Wayne Schmidt's Best Tasting Cake Page

How to make the best tasting cakes from mixes.



For years I'd been dressing up box cake mixes by using milk and soft margarine instead of water and oil to create cakes that tasted richer and had better texture. Then one day I spotted Anne Byrn's book The Cake Mix Doctor and realized the idea was older than I'd imagined. Reading this fascinating book opened my eyes to the fact that while what I'd been doing would indeed help box mixes taste better, I'd been short sighted in realizing all the other ways boxed mixes could be improved. With the intent of creating cakes that were so good that they reduced eaters to masses of quivering gastronomic ecstasy, I embarked on a series of experiments that subjected Mrs. Byrn's techniques to rigorous comparisons to determine which worked best. This web page is a record of those experiments and their results. Both successes and failures are documented with detailed descriptions so readers can decide for themselves which techniques are most likely to satisfy their particular preferences.

Why start with cake mixes? Because the accuracy and consistency of their premeasured ingredients is greater than what most home bakers can achieve. Starting with a boxed mix removes most of the stumbling blocks that ruin many made-from-scratch cakes. Take flour. Depending on how it's scooped a cup can weight anywhere from 120 to 180 grams, one makes a cake that's too wet, the other too dry and tough. Mixes also contain compounds such as emulsifiers that aren't available to home chefs. These chemicals improve consistency, make mixing easier and improving the texture of the cake. They also automatically adjust for minor variations in how the batter is mixed to provide greater reliability in the end cake.

In other words: they're idiot proof.

Many bakers fault boxed mixes for having poor flavor. In many cases they're right, but I've found this is most often the fault of a manufacturer producing a bad formula, not a fault with the concept of a boxed mix. This is proved a few paragraphs down this page with a box mix taste test. Some are terrible, one is outstanding. If you've tried a box mix and were disappointed in the results, the fault might not be with the fact that the cake came out of a box, but rather that it was the wrong box.

My own experience is that if I select the best tasting box mix available and make a few minor changes to the wet ingredients mixed into it the result is a cake that is more consistent, better tasting and has a better texture than any made-from-scratch cake I've made. I believe that anyone following the directions on this web page will be more than satisfied with the resulting cakes. So, without further delay, let's get started.



Many times I recommend specific brands and varieties of cake mixes or frostings. I have not been asked or paid to do so and am not connected in any way with these companies. My recommendations are the result of my independent taste tests.



Test cakes were baked in groups of four, 6-inch pans, each one filled with a one-quarter box recipe. The pans were lined with parchment paper, sprayed with Pam and the outside circled with water soaked Wilton baking strips for even raising and baking. All ingredients were weighed out to within one gram for consistency. Wet and dry ingredients were mixed separately then combined as closely in time as possible. Test batches of all the same ingredients established that this process produces uniform results.

During each step all the test samples are treated identically to reduce variations that are not reflective of changes due strictly to differences in the ingredients.

All cakes were cooled to room temperature (68 degrees F.) before tasting.

My primary criteria are taste and texture... and I don't care how the best of each is achieved. By that I mean that while many bakers advocate the virtues of using natural ingredients such as real butter, I don't care what ingredients are used if they produce the best results. I'll happily switch to butter flavored Crisco or squeeze Parkay margarine if it works better.

When I mention the amount of an ingredient used in making a mix, such as 1-cup extra sugar, this is the amount that would be used in making a complete cake, not the amount used in making one of the 1/4-box test cakes.

In addition to comments addressing how varying ingredients affects taste and texture, I'll also mention how they affect other aspects of cake baking, such as browning, raising, baking time and pan sticking.

So, without further delay, let's get started.


Vanilla Cake:

This plainest of all cakes deserves more respect than it gets. Properly made, a vanilla cake is deliciously rich and satisfying, so good that at its best you may not even want to frost it because all of its complexity of flavor is easily buried under the weight of a super sweet icing.


Mix Brand and Variety Comparison:

Since this web page is about making cakes from boxed mixes, the first step is to find out which brand and variety makes the best cake. To determine this I compared six different boxed white cake mixes.

Each was mixed and baked according to the instructions on the box. The cakes were tested unfrosted because frosting is so powerful that it would overpower both the best and worst tastes. Whole eggs were used because we were looking for the best tasting cake, not the whitest. All of the cakes rose so close to the same height that the differences were insignificant. In spite of displaying varying hues of white on the boxes, they all came out with the same whitish-yellow color. They were all equally moist and tender.

The following list provides the results:

Betty Crocker White: Bland, uninteresting flavor.

Betty Crocker French Vanilla: Very good flavor but not quite as deep as BC's golden vanilla cake.

Pillsbury Classic White: The second most popular choice. Excellent balance between sweetness and depth of flavor.

Duncan Hines Classic White: Foul chemical flavor assaults your taste buds at the first bite. Vile. The batter was extremely thick and difficult to level in the pan. It rose a fraction of an inch higher than the others.

Duncan Hines French Vanilla: Bland initial flavor quickly followed by a baking soda aftertaste. Very bad.

Betty Crocker Golden Vanilla: The richest, deepest flavor of all the cakes. By far the best and better than most made-from-scratch cakes.

I was so surprised at how bad both Duncan Hines offerings were I invited my wife to try them and she agreed that they were terrible. The classic white mix was so bad I'm tempted to obtain a second box from a different store to see if the one we tested had been incorrectly formulated at the factory. It's hard to believe anything that bad would last long in the marketplace when much better mixes are available. Perhaps this is the box mix that gives all other box mixes a bad name. It may be that no one knows how bad it is because it's only served frosted, which masks its foul flavor.

Now that we've identified the best tasting box mix, let's try some experiments to see if it can be improved.

Test 1. Changing the Basic Ingredients:

The instructions on the box say to use water and clear oil. My first 4 experiments to improve this box mix's flavor was to (1) use whole milk instead of water and squeeze Parkay margarine in place of the oil, (2) using melted butter instead of oil and (3) adding two teaspoons of real vanilla to the mix.

Clockwise from upper left: water and oil, milk and squeeze Parkay
milk and melted butter, milk with Parkay and extra real vanilla.

Switching to whole milk and squeeze Parkay was a great improvement, changing the basic cake from something that's nice to eat to a cake that's so good it was almost impossible to stop. Melted butter quickly solidified when it hit the milk, resulting in small lumps in the batter that never completely beat out. However, this had no effect on the resulting flavor. Using butter results in a cake with a better aroma when warm than if oil or Parkay were used, but the flavor was almost indistinguishable from the Parkay cake. If pressed I'd have to say that the Parkay cake had a deeper, richer flavor. I couldn't detect any flavor difference between the cake made with extra vanilla.

What surprised me about this test is that Squeeze Parkay tastes pretty horrible compared to real butter or a good margarine such as Land of Lakes. The only theory I have to explain its success is that the chemicals that make Parkay taste so bad somehow compliment chemicals in the cake mix so that they cancel each other out. Stranger still, when I use Land-of-Lakes margarine, the flavor isn't much different that squeeze Parkay.

The cake made with milk and margarine had a slightly creamier, moister texture than that made with water and oil.

Although the two cakes in the rear appear to be smaller, that's a result of them being further from the camera. In fact all four cakes were the same size and rose the same amount. Differences in the color of the sides of the cake are due to the way in which I used a knife to cut them away from the sides of their pans and do not reflect one being more done than the other.

Conclusion: By all means use whole milk and squeeze Parkay. The resulting cake is wonderfully moist and delicious.


Test 2. Using Buttermilk:

Some of the recipes in The Cake Mix Doctor suggest substituting buttermilk for the water. I tried this using straight buttermilk, buttermilk plus 1/4 cup sugar, 1/12 cup sugar and finally 3/4 cup sugar with the intent of discovering if extra sugar balanced buttermilk's tartness.

Clockwise from the upper left: Buttermilk, buttermilk with 1/4 cup sugar,
buttermilk with 1/2 cup sugar and buttermilk with 3/4 cup sugar

Considering how much stronger buttermilk tastes compared to regular milk, I was surprised that the cake made with buttermilk wasn't more strongly flavored. It didn't have the deep and rich flavor of the cake made with whole milk and had a slight tanginess in the aftertaste that wasn't pleasant. Adding sugar was a disaster. It quickly made the cake overpoweringly sugary sweet.

The texture of the cakes were all the same. However, as sugar increased the cakes browned more attractively, took longer to bake, tended to raise flatter and stuck to the sides of the pans more aggressively.

Conclusion: Using buttermilk in place of whole milk doesn't improve flavor.


Test 3. Adding Sour Cream or Yogurt:

The Cake Mix Doctor also recommends replacing water with one cup of either sour cream or yogurt and one-half cup of milk to compensate for the sour cream and yogurt not having as much liquid as water.

Top left: Sour Cream, Lower left: Sour cream with 1/4 cup sugar
Top right: Yogurt, Lower right: Yogurt with 1/4 cup sugar

As with the buttermilk, I was amazed by how little using either sour cream or yogurt changed the cake's flavor. In the case of sour cream, it tasted most like the cake made with melted butter but with a slightly harsh edge in the aftertaste. The yogurt gave its cake an unpleasant chemical note. Adding sugar once again made the cakes far too sweet.

Although the batters in both cases were thicker than the batters made with milk or water, the resulting cakes rose the same amount. Texturally, they were slightly heavier than a cake made with milk, but the difference was so small that it would be impossible to detect without a side by side comparison.

One interesting point is that the sour cream cake had a batter who's texture was a delight to mix. It's hard to describe but the feel and appearance of the beaters cutting through the batter was great.

Conclusion: Switching to sour cream or yogurt didn't improve either flavor or texture.


Test 4. Adding Pudding or Cream Cheese to the Mix:

Although every box mix I've found states that it already has pudding in the mix so adding more isn't advised, I decided to do so anyway to see what happened. I also tried adding cream cheese as per Mrs. Byrn's recommendations.

Upper Left: One small box of vanilla pudding plus 2 extra cups of milk in one box of cake mix

Lower Left: Vanilla pudding with no extra milk

Upper Right: 1/4 sugar and 1 cup cream cheese plus 1/2 cup milk replacing 1 and 1/4 cup water

Lower Right: 1 cup cream cheese plus 1/2 cup milk replacing 1 and 1/4 cup water


Adding pudding created a slightly denser, creamier texture with an overly strong vanilla flavor. Adding pudding with as much extra milk as called for by the instant pudding directions (2 cups for 1 small box) resulted in a cake that was so moist that after the first bite it dissolved into mush. The milk helped smooth out the overpowering vanilla flavor slightly.

The cakes with cream cheese had a harsh edge, which adding extra sugar did not remove. These cakes also had an unattractive color.

Conclusion: Neither pudding nor cream cheese improved the cake over the basic recipe with milk and margarine


Test 5, Fat Free Milk, Extra Egg yolk, Yellow Cake Mix and Extra Real Vanilla:

Clockwise from upper left: Fat Free Milk, Yellow Box Cake
Extra Egg yolks, Extra Vanilla

Using fat free milk in place of whole milk produced a cake that was only slightly better than one made with water. It appears the fat in the milk plays an important role in both flavor and texture.

While these tests focus on vanilla cake, I was interested in how the most popular box cake of all, yellow, compared. It was slightly yellower, the flavor was not as rich and the texture a little drier.

Adding two extra eggs yolks to a full box mix results in a cake that has a decidedly unpleasant eggy note.

Increasing the extra vanilla to 4 teaspoons per box mix gave a cake that had an overly harsh vanilla flavor and aroma.

Conclusion: Using fat free milk, extra egg yolks and even more vanilla doesn't help.


Additional Tests:

A. I also tried using four whole eggs instead of three per box mix. The flavor didn't change and the cake had a slightly rubber texture.

B. Substituting butter flavored popcorn oil for margarine made a cake with a harsh edge to it and had no improvement in texture.

C. Reducing the time the batter was beaten from two minutes to one had little effect on the cake as did increasing it to three minutes.

D. Reducing the beating time to half a minute resulted in a cake that didn't raise as high and had an overly moist, dense texture. Over beating on high, rather than the recommended medium mixer speed, produced a cake that rose slightly higher, was lighter, but a little too soft.

E. Typical instructions call for cooling the cake in the pan for 10 minutes then removing it to a wire rack to completely cool. During this process much of the cake's moisture is lost. To see what effect different cooling strategies have on texture I cooled cakes in four different ways: as directed, in the pan, in the pan for 10 minutes then covered and finally in the pan and covered as soon as the cake is taken from the oven. As expected, the one covered immediately was the moistest with the other options midway between it and the open cooled cake. What was interesting is that I'd rate it 50-percent moister, not two or three times as moist as I had expected. Evidently the cake's structure is very efficient at retaining moisture even when subjected to air cooling. I suspect the credit for this comes from some of those oddly-named chemicals listed in the ingredients. The covered-right-away cake also had a richer flavor. One possible explanation is that a moister cake's extra water content helps carry dissolved flavors into taste sensors on the palate. One negative to cooling cakes covered is that a stable crust doesn't form, making frosting such cakes difficult. The frosting tends to roll the crust up as the frosting is spread. If a cake is to be eaten unfrosted, definitely cover cakes immediately after removing them from the oven, being careful the cover doesn't touch the top of the cake. You also want to make sure that the cover doesn't curve down in the middle, because that can cause condensation to drip down on the cake and turn it into pudding. I use an upside-down dinner plate. If you wish to frost such a cake, freezing it and frosting it while it's still frozen may help.

F. Using solid or soft margarines improves flavor over using oil, but the overall texture and moistness isn't as good as with squeeze margarine because its natural state at room temperature is a thin liquid.

G. Cake is best eaten as soon as it's cooled. Stored overnight in a plastic bag it loses a little of its richness but is still very good. Freezing in a plastic bag doesn't reduce flavor or texture. I found thawing cake unwrapped preserves most of its moisture. Unwrapped, water vapor initially condenses on the frozen cake making the surface sticky, but by the time the inside of the cake has thawed the surface dries out.

H. Since much of the cake's rich flavor comes from the thin browned top and bottom, I tried making three thin layers and stacking them to maximize flavor. It didn't work. The thin layers lost so much moisture during baking that they ended up like cake jerky.


Overall Conclusions:

1. The best tasting and moistest cake uses whole milk and squeeze Parkay instead of water and oil. No other substitutions or additives made as good a cake.

2. If at all possible, cover the cake immediately after removing it from the oven to retain maximum moistness and flavor.

3. Height wise, box cake mixes are amazingly resilient. Large changes in ingredients and mixing techniques still produce cakes that all rise about the same and have the same texture.

4. While I found that adding ingredients like sour cream, yogurt, real butter, or cream cheese don't improve flavor or texture, they may nonetheless serve a physiological purpose. Using them makes the baker feel as if she or he is doing something special that will make their cake better. Since half the fun in making cake is the mixing and baking, using such ingredients may serve to make the overall experience more rewarding (even if the resulting cake isn't as good.)

5. In almost all the cases the differences between the tests was so small that unless you are an expert taster or doing side-by-side comparisons you're probably not going to notice much difference. The one exception, to my mind at least, is the simple change of using whole milk and squeeze margarine in place of water and oil. This makes a cake that outshines all the others.

6. Starting with Betty Crocker Golden Vanilla cake mix and using whole milk and squeeze margarine produces a cake of such deep, rich, satisfying flavor that it doesn't need to be frosted, particularly if it is covered to cool immediately after being removed from the oven.

7. The top of the cake should be evenly golden brown to insure that the bottom is also baked golden brown. The caramelization that creates the browning is an important factor in developing this cake's rich flavor.

8. For the ultimate in flavor, time the cake to be done one-half hour before eating. Covered and with the baking strip left on, it should still be warm, greatly improving the moistness and depth of the flavor.


A Few Words About Frosting:

Even the thinnest layer of frosting will overpower the flavor of the cake so much that it makes little difference which box mix is used. Additionally, frosting is so sweet that it quickly becomes too much of a good thing... and this is from a man with several demanding sweet tooths. Consequently I recommend that people truly interested in good tasting cakes try eating them without frosting. The depth and range of flavors is much greater. However, if you have to have frosting by all means use the best.

I've tried a dozen different vanilla frostings and none tasted very good. The best I ever tasted was Pillsbury's outstanding French Vanilla Frosting... which they discontinued in 2007. Comparing the remaining major brands provided the following results:


Betty Crocker's was very thin, almost liquid and tended to droop off the sides of cakes. It was excessively sweet. Pillsbury's was slightly thicker and not quite as sweet. Both of these frostings had a fine grittiness. Duncan Hines Creamy Home-Style was the smoothest and although it had the firmest texture it was the easiest to spread. Most importantly, it was by far the best tasting frosting.

While Duncan Hines was the clear winner, it still doesn't come close the matching the depth of flavor of Pillsbury's French Vanilla frosting.

The Duncan Hines can be improved by adding one drop of yellow liquid food coloring, 1 and 1/2 teaspoons of imitation butter flavor and 1 teaspoon of real vanilla extract. This still isn't as good as Pillsbury's French Vanilla but it's clearly better than the others. The additions make it a little thin but very creamy. This makes it a little too soft for fine decorating. Adding powdered sugar firms it up. Still, it has a sharp edge to the aftertaste, which I was able to determine came from the artificial vanilla in the original frosting.

After completing the test above, I discovered Betty Crocker released a new frosting named "Butter Cream." I tried it and wasn't impressed. It tasted more like sweetened crisco than anything else and dispite its name, didn't taste like butter at all.

I decided to experiment with homemade frostings. Here's what I found:

Basic Buttercream:

Most recipes call for 1 stick (1/4-pound) of butter (I prefer salted), 220 grams (half a box) of powdered sugar, up to three tablespoons of whole milk and 2 teaspoons of vanilla. I tried several variations and found that they all came out too sweet, too buttery and ended with a harsh edge. The edge I traced to the vanilla, even though I was using top quality Madagascar Bourbon Vanilla Extract. I found that reducing the butter to 6 tablespoons (3/4 stick) produced a better sweet-to-butter balance. Cutting the vanilla all the way down to 1/4 teaspoon eliminated the harsh vanilla aftertaste yet provided enough vanilla to enrichen the overall flavor and mellow out the sugar impact. Finally, adding 1/4 teaspoon popcorn salt, which is finer than table salt and blends better, also helped deepen the flavor.

This is a very good tasting frosting that has a rich, deep, mellow flavor. Less milk can be used to make a stiffer frosting, but it will end up being gritty because there isn't enough liquid in it to soften the sugar.

Italian Buttercreams:

These use boiling sugar syrup drizzled into beaten egg whites to cook them and create a light, smooth frosting. I tested two different recipes and found them both to be too thin and had an eggy flavor that didn't appeal to me.

French Buttercreams:

These are made similar to Italian buttercreams except they used whole eggs or egg yolks beaten into a froth before the sugar syrup is drizzled into them while beating. I tested one recipe using whole eggs and one using only egg yolks. Both tasted rich and weren't too sweet. But, because of the egg yolks they came out so yellow that coloring them would be difficult. Additionally, they were so thin they'd sag on a cake and also had a slight eggy taste.

Cooked-Flour Frostings:

These use a pudding made of three tablespoons of flour mixed into a cup of milk and cooked in a saucepan until thick. Cooled, then added to a buttercream frosting they stretch out the sugariness of the buttercream so the resulting frosting isn't overly sweet. The problem I had with several attempts is that no matter how long I beat the frosting, it would invariably start to separate and get grainy when stored in a refrigerator for just a few hours. The texture was a little gelatinous when fresh and the flour imparted an odd flavor that didn't seem to belong. I tried cake, all purpose and unbleached flours. They didn't make any significant difference to the flavor or texture.

Marshmallow Cream Buttercreams:

Marshmallow cream is already so sweet no additional sugar is needed to turn it into a frosting. After several experiments that best result I achieved used one jar of marshmallow cream (7 ounces), 3/4-stick of butter and one tablespoon of whole milk beaten together. I still consider it a failure. The initial texture feels like you're chewing crisco and the strong hint of marshmallow seems out of place.

Revisiting Duncan Hines Homestyle Vanilla Frosting:

While looking for an easier way make a buttercream frosting, I began experimenting again with Duncan Hines frosting. I discovered that by simply beating 1 and 1/2 sticks of room temperature salted butter into it at the highest speed my mixer allows, I get a rich, mellow, light frosting that's sweet enough to be satisfying and holds its shape well enough to permit simple decorating. Some people claim it's a little too buttery and in large quantities I agree. In those cases I recommend one stick of butter per 16-ounce tub of frosting. This simple frosting is much easier to make than any from-scratch frosting, tastes better than any homemade frostings I've made including the buttercream recipe mentioned above. An additional benefit is that it doesn't form a crust after being exposed to the air for up to six hours. The surface remains smooth and creamy.


One word of caution: like all butter-based frostings, this one will soften as the room temperature goes up, which could lead to the frosting sagging on the cake.



Lemon Cakes:

Unlike vanilla cakes, citrusy cakes like lemon cake have enough flavor to stand up to the sweetness of frosting. One of my favorites is lemon cake. I compared the three available brands made with both water and oil as directed and with whole milk and squeeze margarine. The best tasting cake was Betty Crocker with a bright, clean lemony flavor. Duncan Hines was second with a very weak lemon taste. Pillsbury was a distant last with a harsh chemical note. Unlike vanilla cake, these taste better and are lighter if made with water and oil rather than milk and margarine.

I substituted 1/4 cup of lemon juice for an equal amount of water and found that the resulting cake was slightly too lemony-tart when eaten unfrosted. BUT! After frosting, the sweetness of the frosting was perfectly balanced by the now more acidic cake to produce something so delicious than half an hour after eating it my palate was still fresh, clean and had a pleasantly lingering sweet lemon taste.


Lemon Frosting:

Betty Crocker was the best tasting lemon frosting with a bright, lemon flavor and smooth texture. (It's so good we keep some of it in the refrigerator for when we feel like a taste of something sweet and zesty.) Pillsbury was a close second with a grainier texture and a weaker lemon taste, particularly in the aftertaste. Duncan Hines was very poor with practically no lemon flavor.


Using the two Crocker products, adding lemon juice to the mix and cooling the cake covered before frosting yields a moist, zesty lemon cake people will rave about.



Chocolate Cakes:

As with vanilla cakes, the first step was to conduct a thorough taste test of all available brands and varieties. Combing five different supermarkets turned up 17 mixes:

Baking each according to the instructions, except to cool them covered for maximum moistness, I cut sample slices to compare appearance...

From left to right, starting with the top row:

Betty Crocker Butter Recipe, Betty Crocker Dark, Betty Crocker Chocolate Fudge, Betty Crocker Triple Chocolate Fudge, Betty Crocker German

Duncan Hines Devil's Food, Betty Crocker Devil's Food, Duncan Hines Dark Chocolate Fudge, Pillsbury Dark Chocolate, Duncan Hines Red Velvet

Betty Crocker Milk Chocolate, Pillsbury Devil's Food, Duncan Hine's Swiss, Pillsbury German, Duncan Hine's German

Betty Crocker Decadent Supreme, Duncan Hines Decadent Triple Chocolate


... and then subjected them to taste comparisons. Before continuing I need to explain something about the difference between eating versus tasting cake, chocolate cake in particular.

Watch most people eat chocolate cake and you'll notice they chew a bite of cake once or twice then swallow it whole. They really aren't tasting it. They "notice" that they're eating chocolate cake but seldom take the time to fully savor what they are eating. This is partially the fault of our fast-paced western culture, which drives us to complete each task as quickly as possible so we can move on to the next sensation. It's also partially the product of an almost world wide mind set that chocolate is always good and more is always better.

When applied to chocolate cake these factors drive us to rush through our slice of cake without being fully aware of what we're tasting because "if it's chocolate it must be good."

Prepare yourself for a painful fact of life: just because something has chocolate in it doesn't make it good.

Not all chocolates taste good. Take Hershey's. Sometimes it's tasty but most of the time it's disgustingly harsh. (For an in-depth review of Hershey's quality, please take a look at the HERSHEY'S VINTAGE PAGE.) The same is true of all chocolate brands. Dump bad chocolate into a mix and you're going to end up with a cake that tastes terrible.

Contrary to popular trends, more chocolate is not always better. For optimum flavor, chocolate cake requires a careful balance between the sweetness of sugar and chocolate's natural bitterness. Too much of one or the other and the result is inferior. In the case of chocolate, too much and the cake will have a harsh, edgy flavor which, if someone takes the time to consider what they are eating, isn't very pleasant.

For my taste tests I carefully analyzed what I was tasting: was it too sweet, sour, or edgy? Was the aftertaste pleasant or did it quickly turn sour? Most important of all: did I enjoy the sensations experienced while eating it? To my surprise, the answer to this last question was "no" with the vast majority of cakes.

For the actual taste test, I number coded each cake and scrambled the order of the cakes so I wouldn't know what brand or variety I was tasting. Cakes were tested in three groups based on color: light, medium and dark. The best tasting cakes from each group were then compared to determine the cake with the best overall flavor. The following list presents what I found:

Light Colored Cakes:

Betty Crocker German Chocolate: Smooth, Mild, pure chocolate flavor with no edginess. THE WINNER IN THIS GROUP.

Betty Crocker Milk Chocolate: Moister than BC's German but it had a slight edginess or harshness that distracted from the chocolate flavor.

Pillsbury's German Chocolate: The flavor was similar to Betty Crocker's Milk Chocolate cake but wasn't as moist.

Duncan Hines German Chocolate: The same as Pillsbury's German Chocolate. Side by side I couldn't tell them apart.

Medium-Dark Colored Cakes:

Betty Crocker Butter Recipe Chocolate: Similar smoothness of flavor to Betty Crocker's German Chocolate but with a little too much chocolate so that it ended up tasting slightly harsh.

Betty Crocker Dark Chocolate: Slightly harsher than the Butter Recipe mix.

Betty Crocker Chocolate Fudge: Identical to BC's Dark Chocolate mix. (I wouldn't be surprise to learn that all three were the same, the only difference being that the Butter Recipe mix calls for butter instead of oil.)

Betty Crocker Triple Chocolate Fudge: Very harsh with a sharp, unpleasant edge.

Duncan Hines Devil's Food: Very moist but sharp and bitter.

Pillsbury Devil's Food: Slightly sweeter than the others but still had a sold chocolate flavor. Very good. THE WINNER IN THIS GROUP.

Duncan Hines Swiss Chocolate: This was an odd one because it had a very weak chocolate flavor yet it tasted edgy like it had too much chocolate in it.

Duncan Hines Red Velvet: Smooth and mellow chocolate taste but with an unpleasant chemical note that I assume was from the red food coloring.


Dark Chocolate Cakes:

Duncan Hines Dark Chocolate Fudge: Very moist. Strong chocolate flavor with a deep richness. I found it good but overpowering.

Pillsbury Dark Chocolate: Odd, gray hue to its color. Smooth, solid, pure chocolate flavor. THE WINNER IN THIS GROUP.

Betty Crocker Decadent Supreme: Very smooth, chocolaty flavor, but with a harsh edge.

Duncan Hines Decadent Triple Chocolate: Super moist - more like a very light brownie than a cake. So moist it was impossible to cut cleanly. Strong, bitter, chocolate edge.


Overall Winner:

My taste preference is for Pillsbury's Dark Chocolate cake with Betty Crocker's German Chocolate cake coming in a close second. BC's German Chocolate cake is slightly sweeter and mellower, but the chocolate flavor is a little weak in a head-to-head comparison.



I switched to milk and squeeze Parkay and as with the vanilla cake these substitutions significantly improved flavor and texture. I did not repeat all the other experiments because considering the overall negative effect they had on vanilla cake I decided they would probably not improve these cakes.

Adding 1/4 cup of sugar to the dark chocolate mix didn't sweeten it but did mellow the slight edge to the aftertaste. I count this as an improvement. Adding 1/2-cup of sugar mellowed it a little further and introduced a pleasant sweetness. This is the formula I'll be using from now on. The added sugar slows baking so it takes two more minutes to be done. Another benefit to adding 1/2-cup of sugar is that the cake comes out perfectly flat. Even with baking strips, mixes made without the extra sugar end up with a high center.

I also experimented with a 50-50 mix of the dark chocolate and German chocolate mixes. It was a complete failure. Rather than combining the best of both mixes it brought out the worse of them. The resulting cake tasted like I couldn't make up what it wanted to be: sharp or mild, strong or sweet.


Chocolate Frostings:

I tasted every chocolate frosting I could find and was disappointed with all of them. They were universally too sweet and so chocolaty that they tasted harsh. The best I found was Betty Crocker's Milk Chocolate frosting, but this was the winner not because it was good but because all the other were so bad. All the frostings also suffered from being too thick. Spreading them on moist chocolate cakes, which seem to be more delicate than vanilla cakes, invariably resulted in the top layer of the cake being torn off.

My problem with chocolate cakes is that because chocolate frostings are so overpoweringly sweet and chocolaty, the cake becomes little more than a delivery system for the frosting (to quote Alton Brown of the Good Eats television show.) What's needed is a smooth, light, whipped frosting with a mild chocolate flavor that lets the cake's flavor come through.

I decided to employ what I discovered while experimenting with vanilla frostings to "fix" the Betty Crocker frosting. After many failures, I found I could create a frosting that tasted like chocolate but without the harsh edge of all of the chocolate frosting I'd tried. Starting with one tub of Betty Crocker Milk Chocolate frosting, I beat into it 3 and 1/2 sticks of room temperature salted butter, 320 grams of powdered sugar, then 4 tablespoons of whole milk, added one at a time while beating. The resulting frosting has a light chocolate color, isn't overly buttery, is light textured yet firm enough for simple decorating and has a solid, non-harsh chocolate flavor that complements chocolate cake rather than over powering it. Like the buttercream frosting mentioned earlier, this chocolate frosting doesn't form a crust after being exposed to the air for up to six hours. The surface remains smooth and creamy.

One word of caution: like all butter-based frostings, this one will soften as the room temperature goes up, which could lead to the frosting sagging on the cake.



Several years ago I conducted a chocolate cake test where I compared many different types of cakes: everything from light sponges to the internationally famous Sacher Torte. What I discovered was that in terms of flavor and moistness, the overall winner was the plain old American brownie. No other cake matched it for texture or flavor. It's unique in that it is one of the few cakes with enough flavor to stand up to even the sweetest of frostings, whether they be vanilla or chocolate. The following list presents the results of my taste test:

Pillsbury Fudge Brownie: Used to be good, but in 2006 they changed the formula and the current version is very bad.
Pillsbury Chocolate Chunk: chocolate flavor was too harsh
Pillsbury Double Chocolate: taste too close to semi-sweet chocolate

Betty Crocker Fudge Brownie: weak flavor
Betty Crocker Chocolate Chunk: bitter
Betty Crocker Original Supreme: very good THE WINNER

Duncan Hines Double Fudge: sharp, chemical taste
Duncan Hines Chewy Fudge: not sweet enough
Duncan Hines Chocolate Chunk: chocolate too strong

Albertson's: bitter

New Organics: flavor too close to semi-sweet

Ghirardelli: sharp, over-strong semisweet taste


Betty Crocker Original Supreme was the overall winner. Experimentation showed that leaving out the chocolate syrup included in the box improved flavor by creating a brownie that has a more balanced sweet-to-chocolate flavor. Using whole milk and squeeze Parkay in place of water and oil also helps.

The only problem is that a brownie is too moist and heavy for a layer cake. They also carry with them an informal feel that prevents them from receiving the respect they deserve.


Strawberry Cakes:

Pillsbury strawberry cake came out pink and tasted mostly sweet. Duncan Hines had a strange salmon color but tasted slightly more like strawberries.

Both mixes had a chemical note and didn't taste anything like real strawberries. Adding artificial strawberry flavoring made things worse. I'm assuming that "strawberry" is one of those flavors that have to be real and fresh to be good. This raises the problem that most commercially grown strawberries are from varieties that look good, store well and are tough enough to ship... not for flavor. Worse still, they are picked before being completely ripe and haven't been allowed to develop their full flavor. Used in a cake they might not do much better than artificial flavorings. Strawberry frostings were even worse.

I hope to solve this problem by growing Albion strawberries (a new, multiple award winning variety for flavor) this summer and puree them into a plain white cake mix. Please check back to see how this experiment turns out.


Zesty Lime Cake:

Start with a Betty Crocker White Cake mix and mix as directed with oil and water, replacing 3 tablespoons of the water with three tablespoons of lime juice. I prefer to not use milk and squeeze margarine in this recipe because they give the cake a heavy flavor that detracts from the brightness of the lime juice. Bake and cool covered according to the directions.

For the frosting, beat three sticks (12-ounces) of softened butter, 6 ounces of cream cheese, 3/4 box (12 ounces) of powdered sugar until light a fluffy. Slowly add two tablespoons of lime juice as you continue to beat.

The resulting cake has a bright, clean lime flavor that's a real treat. The frosting is very zesty with just enough sugar to balance the cheese and lime juice's acidity.

I prefer to leave the cake uncolored to help draw attention to the light green frosting. Lime frosting is tricky to color properly. There are two accepted hues: a very pale green corresponding to a lime's internal color or a dark green reflecting it's external color. I think going light creates a more refreshing, cooler appearance. The problem is that it's very easy to add too much green food color and end up with something that looks more like mint frosting than lime. To avoid this I dip just the tip of a toothpick in liquid green food color to pick up the smallest possible amount. Drawing the toothpick through the frosting to wipe off the color makes it easy to work your way up slowly to the desired hue.



Unlike white and chocolate cakes, which had many varieties, I could only find two makes of spice cake mixes. Betty Crocker was the clear winner with a good balance of sweetness and sprightly spiciness. Duncan Hines was bland in comparison. Flavor and texture are not improved by substituting milk and squeeze margarine for water and oil because their richness dulls the brightness of the spices.

Spice cakes demands cream cheese frosting. The following taste test will help you get the best brand:


Cream Cheese Frostings:

Cream Cheese frosting is one of the great flavors of all time. The question is: "Who makes the best tasting frosting?" To answer this I purchased all the varieties I could find and compared them. Here are the results:


Pillsbury...,,,,biege....average..chemical edge....gritty, sharp cream cheese aftertaste

Stator Bros..white.....firmest..strongest, sweetest...smoothest texture, richest flavor

Whipped...,,white........firm........no aroma....weak cream cheese taste, harsh aftertaste

Homestyle . white.......soft.....chemical edge..............unpleasant musky flavor

Crocker...,,,white......softest.....no aroma.................sharp chemical aftertaste

For me the winner was Stator Bros. However, none of these was really very good. For an easy to make and vastly tastier frosting simply whip together 1 cup (2 sticks) of room temperature butter, 6-ounces of room temperature cream cheese and one box (454 grams of powdered sugar. This makes a frosting that's smoother and far richer than any of the brands tested. If you want the frosting to have a lighter texture, add 1 to 3 tablespoons of whole milk to it one at a time, whipping thoroughly after each addition.


Mint Cake:

I couldn't find a mint flavored cake mix so I started with a plain white cake and experimented with different additions to come up with the best tasting mint cake I could. I discovered that peppermint extract gives a sweeter, brighter flavor than mint extract, which tasted a little peppery. Also, using water, oil and egg whites allows the peppermint flavor to come through better than a cake made with milk, margarine and whole eggs. The cleanest flavor came from adding 2 tablespoons of peppermint extract to a full box mix. Any more and the cake developed an aftertaste that burned a little.

I also could not find a peppermint frosting so I made one using one tub (16 ounces) of Duncan Hines Homestyle Vanilla Frosting into which 1 and 1/2 sticks of soft butter and 1 and 1/2 teaspoons of peppermint extract had been added. As with the cake, adding more extract becomes overpoweringly harsh. Adding one drop of green liquid food coloring gives the frosting a very light green tint that works well with the mint. Leaving the cake white hightens the effect more than tinting the cake green.



Ever had a cake come out too dry? How about having one fall during cooling because it was underdone or too moist? The source for most of these problems is inconsistency in the amount of ingredients used. My solution is to weigh everything out to the gram using an electronic scale. For a standard vanilla or chocolate boxed cake mix weighing 520 grams (although the boxes usually state the net weight is 517 grams, I find they are always a few grams heavier) I mix in 300 grams of whole milk, 76 grams of squeeze Parkay and 144 grams of beaten egg. The mix is always beaten the same amount, 30 seconds on low, scraped, then 2 minutes on medium. I keep a log of how long to bake each cake depending on the pan size: 16 minutes for medium cupcakes (always precisely measured scoops), 24-minutes for 1/4 of a boxed mix in a 6-inch pan, 23-minutes for 1/2 of a mix in a 9-inch pan and so on. I always used thoroughly soaked baking strips and clean aluminum pans lined with parchment paper. Consistency is the key to reliable results.


Putting The Recipes To The Test:

I baked chocolate, vanilla, lemon, spice and mint cupcakes using the recipes on this page and took them to work to see how other people rated them.

The rapidity with which they disappeared and rave reviews proved that these cake recipes are enjoyed by everyone. I also used the opportunity to determine which flavors were preferred. Chocolate was most favored, vanishing twice as fast as any of the others. Lemon, spice and mint tied for second. Vanilla came in last, though still appreciated by everyone who tried them. The only complaint was that there needed to be many more of each flavor. In a follow-up test, I baked five dozen chocolate cups cakes and frosted half of them with chocolate frosting and the rest with vanilla frosting. The cupcakes with vanilla frosting were twice as popular as those with chocolate frosting.


A Word of Warning:

Box mix manufacturers are continually experimenting with their recipes to make them taste better or be cheaper to produce. Because of this a brand and variety that tastes great in 2010 may be changed to something disgusting in 2011. If your cakes suddenly don't taste like they used to, the fault may be with the manufacturer. If that happens, it may be necessary to retest all the available cake mixes and find a new best-tasting variety.



Return to main page to browse over 100 other subjects: everything from metal detectors to the strange world of lucid dreaming.