The Science of Biscuits

Butter or shortening? Milk or buttermilk? What is the true key to flakiness? These are the questions that plague biscuit makers everywhere, and they are the questions that I have set out, over the past 13 months, to answer once and for all.

When you ask how to make biscuits, it seems that everyone and their grandmother (especially their grandmother) has an opinion. If you look up a hundred biscuit recipes, you'll get some commonalities... and a lot of specific, mysterious "key" steps that sound more like potion-making instructions from a Harry Potter book than a real recipe.

Does the amount of salt really matter down to the 1/3rd of a teaspoon? Is it really key to knead exactly 15 times, or to use a baking mat, or to use buttermilk? Or are some of the mysterious steps in these recipes as arbitrary and nonsensical as cutting off both ends of the ham?

There's only one way to find out: experimenting. For the past year, I've performed dozens of biscuit experiments. I've tried to be as scientific as possible by:

• changing only one thing at a time (the amount of fat or the type of liquid, for example)
• doing blind taste tests with friends (i.e. not explaining the difference in recipes until after I get opinions).
• testing multiple types of biscuits on the same day (so that taste testers can say "biscuit A is better than biscuit B" rather than having to say, "Yeah, I think these taste better than the ones you made last week!")

In this instructable, I will show the results of these experiments and my results for the "ultimate" biscuit recipe. But first a little about technique: how do you make biscuits in the first place?

Making biscuits is basically composed of seven steps:

1. Mix some dry ingredients
2. "Cut" in some fat
3. Mix in some liquid
4. Knead the dough
5. Roll out the dough
6. Cut biscuits
7. Bake

Most of these steps are pretty straightforward, but here are some tips and tricks I've learned after making about a zillion biscuits:

1. Mix some dry ingredients
   • No tips needed. Just mix them with a fork (or anything else handy).
2. "Cut" in some fat
   
   • This can be the hardest step. The goal is to turn your fat into a bunch of little pea-sized pieces coated with the dry ingredients.
   • Preferred method: If you're using butter, it's easy: just grate the butter with a cheese grater.

   

   • Other methods include:
     • Squishing the fat around with a fork until it's a bunch of little pieces coated in flour.
     • Blending the fat in with a food processor.
     • Using two table knives (or scissors) to cut the fat up until it's a bunch of little pieces coated in flour.
     • Using a pastry cutter (not much better than a fork)
     • Using your hands
3. Mix in some liquid
   • Again, just mix with a fork.
4. Knead the dough
   • Fold the dough in half. Squish it. Repeat about 10 times.
5. Roll out the dough
   • Roll with a pin or pat with your hands until it's about one inch thick.
6. Cut the biscuits
   • Traditionally, you use a round biscuit cutter. I like to use a knife and cut squares, so I don't have left-over pieces to re-knead and re-roll.

1. Bake
   • Put the biscuits on a baking sheet and stick them in the oven.

And now, on to the experiments!

Baking powder is a field of contention. Some people think the amount of baking powder used is very specific and important, while others seem to think it's almost unnecessary. I've even read one book that said that rising occurs because of the butter in the biscuits, not the baking powder.

So what difference does baking powder really make?

To find out, I made five different batches of biscuits. The only difference between the batches was the amount of baking powder used:

• Batch 1: 1/2 teaspoon baking powder
• Batch 2: 1 teaspoon baking powder
• Batch 3: 1.5 teaspoons (1/2 tablespoon) baking powder
• Batch 4: 3 teaspoons (1 tablespoon) baking powder
• Batch 5: 6 teaspoons (2 tablespoons) baking powder

I then took a time-lapse video of the biscuits as they baked. The pan has three representative biscuits from each batch. Oven temperature is 425 F.

The video shows pretty clearly that a lot more rising occurs when you use more baking powder. This picture, taken right after the biscuits had finished baking, shows the same thing:

But what about taste?

Both taste testers (my wife and I) thought that batches 1 and 2 (with only a tiny amount of baking powder) were noticeably denser and moister than the others. The texture wasn't the nice, flaky, classic biscuit texture we both like.

For batches 3, 4 and 5 (with half to double the base recommended baking powder amount), the taste difference was almost indistinguishable. My wife thought the batch with the most baking powder was best, while I thought it tasted a tad too salty and chemically. My favorite was the middle batch, with 1.5 teaspoons of baking powder, but my wife thought this one was a little too dense.

Conclusion: More baking powder makes the biscuit rise more (imagine that!). About 1 tablespoon of baking powder per 2 cups of flour seems to be about the right amount, but even halving or doubling this amount should not ruin your biscuits.

A big variation in biscuit recipes is the amount of liquid (milk, water, buttermilk) that the recipe calls for. Some biscuits, like drop biscuits, call for a lot of liquid, so the dough is super sticky and you have to "drop" lumps of it onto the pan. Others call for very little liquid, so it's difficult to mix the dough without over-kneading it. What is best? How does the amount of liquid affect the final product?

To find out, I made five batches of biscuits that were identical except for the amount of milk used. Each batch uses two cups of flour and normally makes 8 biscuits, but I only cooked two biscuits from each batch so that I could make sure the baking conditions were exactly the same for all of the biscuits.

• Batch 1: 1/2 cup of milk (1:4 ratio)
• Batch 2: 3/4 cup of milk (3:8 ratio)
• Batch 3: 1 cup of milk (1:2 ratio)
• Batch 4: 1-1/4 cups of milk (5:8 ratio)
• Batch 5: 1-1/2 cups of milk (3:4 ratio)

I then took a time-lapse video of the biscuits as they baked. The pan has two representative biscuits from each batch. Oven temperature is 425 F.

The video shows that you get a lot more rising for drier biscuits. You can see this well in the picture of the final biscuits as well:

How about taste?

Surprisingly, all of the biscuits were extremely good. I went into this thinking that the amount of liquid ratio was probably the most important variable in the whole recipe. It turns out that it's not. The very dry biscuits were really good, as were the very wet "drop" biscuits. My favorites were probably the ones in the middle (with 1 cup of liquid per 2 cups of flour), but that was by a very narrow margin, and my decision may have been influenced by the fact that a 1:2 ratio is easy to remember.

What about handling the dough?

Again, the 1:2 liquid:flour ratio was about the easiest dough to handle. It was slightly sticky, so you could save some mess by going with a 3:8 liquid:flour ratio. The very dry biscuits were difficult to get mixed, and the very wet biscuits were, well, very wet.

Conclusion:

1 cup of milk (or water, or buttermilk) per 2 cups of flour is a pretty good ratio, but it's really not that important.

Side note:

This experiment also (mostly) debunks the theory that you have to make your biscuits by using a scale rather than using cups and teaspoons. Yes, it's more exact to measure the actual mass (or weight) of your ingredients than it is to measure volume, but this experiment shows that huge variations in the amount of flour per liquid can still result in good biscuits. The small variation in flour amount you'll get by measuring flour by volume will not ruin your liquid to flour ratio enough to ruin your biscuits.

Side note on the Side note:

I say mostly debunks because, theoretically, some other ratio could be super crucial to the science of biscuit making. We've already seen that the baking powder : flour ratio and the liquid:flour ratio have a lot of leeway, though, so unless the fat : flour ratio needs to be honed to perfection, a small measurement error just isn't going to make much of a difference.

Another big variable in biscuit recipes is the amount of fat (butter, shortening, etc.) that the recipe calls for. Usually its in the ballpark of 1/3 or 1/2 cup per 2 cups of flour, which is about half the amount of fat per flour that you put into a pie crust.

But what difference does the fat make? Are biscuits fluffier with more or less fat? Richer? Tastier?

To find out, I made three batches of biscuits that were identical except for the amount of fat used (in this case, butter). Each batch uses two cups of flour and normally makes 8 biscuits, but I only cooked two biscuits from each batch so that I could make sure the baking conditions were exactly the same for all of the biscuits.

• Batch 1: 1/4 cup of butter (1:8 ratio)
• Batch 2: 1/2 cup of butter (1:4 ratio)
• Batch 3: 1 cup of butter (1:2 ratio)

I then took a time-lapse video of the biscuits as they baked. The pan has two representative biscuits from each batch. Oven temperature is 425 F.

As we saw with the liquid experiment, "drier" biscuits tend to rise more (in this case, the biscuits with less fat tend to rise more). in this case, it appears that the biscuit structure is just a lot more stable (structurally speaking) when there's less butter. When you get a lot of butter, you're kind of filling your biscuit with holes, which makes it unable to bear its own weight to rise very far.

So, you definitely get flatter biscuits when you add more butter (incidentally, the same is true with cookies... if your cookies are running all over the pan when they bake, just reduce the amount of butter/shortening and they'll hold their shape much better).

But what about taste?

This was a tough one. Increasing the amount of butter definitely makes the biscuit "taste" softer, more crumbly, and more flaky. I usually associate flakiness and softness with size; you expect a big biscuit to be fluffy and soft, and a biscuit that doesn't rise to be dense. But, at least right out of the oven, the super buttery biscuits really melt in your mouth.

But the biscuits with 1 cup of butter seemed too rich to my wife and I, and I suspect that after cooling down, they would be dense and gross. The biscuit with very little butter wasn't as soft and flaky, but it was more structurally sound.

Conclusion:

None of these biscuits were bad, but again, the middle option seems to be the best choice. A ratio of 1/2 cup of butter to 2 cups of flour (1:4) seems to work pretty well.

Some biscuit recipes call for butter. Others demand shortening. Still others swear by cream cheese. Some even claim a mix: I read one article that claimed that shortening led to better flakiness and butter led to better taste, so an even mix of both would lead to the ultimate biscuit.

We've already seen what the amount of fat does to a biscuit, but what does the type of fat do?

To find out, I made three batches of biscuits that were identical except for the type of fat used. I was cooking for a crowd, so I cooked each batch separately, one after the other, and asked my guests to let me know which batch tasted the best. I did not tell my guests what the difference in the batches was until everyone had told me their opinion.

• Batch 1: shortening
• Batch 2: butter
• Batch 3: cream cheese

Judging by appearance, I could not tell the shortening biscuits apart from the butter biscuits, but the cream cheese biscuits were a little cakier and less flaky looking.

Taste-Test Results:

• Four of eight people could tell no difference in taste between the shortening and butter biscuits.
• Two of eight people thought the shortening biscuits tasted better than the butter biscuits.
• One of eight people thought the butter biscuits tasted slightly saltier, but not necessarily better.
• One of eight people thought the butter biscuits tasted slightly saltier and slightly better.
• The cream cheese biscuits were pretty unanimously agreed to be blander and cakier.

Note:

I (and my taste-testing friends) were pretty surprised by these results, so I did similar experiments about 5 times. Every time, the results were a wash... butter and shortening seem to be very difficult to distinguish in biscuits.

Conclusion:

Either butter or shortening is fine; there's no clear advantage of one over the other (except that butter is easier to incorporate into the dough by hand, especially if you grate it like cheese). Cream cheese is a decent (but inferior) alternative if you're worried about fat content (cream cheese has about half as many calories from fat as butter or shortening). Cream cheese might be an especially good option if you're making biscuits and gravy (where the biscuit is more of a vessel for the gravy anyway).

All biscuit masters agree: if you're going to use butter, use real butter, not that fake crap (margarine).

But margarine is usually several times cheaper than butter, and some argue that it's healthier. Could it really be all that bad for making biscuits?

To find out, I made two batches of biscuits that were identical except for the type of fat used.

• Batch 1: margarine
• Batch 2: butter

I cooked both batches on the same cookie sheet at 425 F, so baking conditions were identical. I had six taste-testers, and I did not tell them which biscuits were which until they'd given their opinions.

It was difficult to tell any difference in appearance between the biscuits. Both turned out pretty nicely: fluffy, flaky, and crispy on the thin outer crust.

What did the taste testers say?

• 5 of 6 taste-testers preferred the margarine biscuits

This surprised me. I asked how sure everybody was of the difference? Three of the five that chose margarine biscuits were only 20% sure of their choice. One was 80% sure, and the last (a smart alec) was 64% sure.

What reasons did the un-witting margarine lovers give? Better texture and taste.

What about the dissenter, who liked the butter biscuits better? She was also only 20% sure, and said that the butter biscuits (which she did not know were butter biscuits) were moister.

Conclusion:

If you're a purist and just can't stomach the "fake crap", then by all means use butter. But if you want to save a few bucks and a few calories, margarine is a great alternative that may even make the biscuits taste better.

Almost everyone, if you ask them, will tell you that buttermilk makes better biscuits. But does it?

I did this experiment many times. Sometimes I used "real" store-bought buttermilk. In this first experiment, I didn't have any store-bought buttermilk handy, so I made my own:

• Batch 1: Buttermilk (2% milk plus 1 tablespoon of lemon juice, let stand for 7 minutes)
• Batch 2: 2% Milk

Baked at 425 F. Four taste-testers.

Taste Testing Results:

• Two of four couldn't tell a difference.
• One of four voted for the buttermilk biscuit (because of "better taste")
• One of four voted for the milk biscuit (because, "the other one has better taste, but this one has better texture")

Conclusion:

Inconclusive. Buttermilk isn't blow-everyone-away better, but it might be a little better. More data is needed (read on).

To solve the milk vs buttermilk question, I decided to make two giant batches for a bunch of guests. I didn't tell anyone what the difference between biscuits was until after they'd made their choice.

• Batch 1: Buttermilk (half was "real" store-bought buttermilk, half was 2% milk plus lemon juice)
• Batch 2: 2% Milk

Baked at 425 F. Eight taste-testers.

Taste-Testing Results:

• Two of eight people thought the milk biscuits were maybe a little "saltier", but couldn't decide which they liked better.
• Six of eight people could not tell a difference.

Conclusion:

Buttermilk doesn't seem to make a difference.

So if buttermilk isn't really better than milk, that begs a question: could you just use water instead of milk? Would taste-testers be able to tell the difference?

To find out, I made two batches:

• Batch 1: milk
• Batch 2: water

Both were baked at 425 F. There were five taste testers (who did not know what was in the biscuits they were tasting):

Taste-testing results:

• Three of five thought the water biscuits were fluffier and better overall.
• Two of five thought the milk biscuits had a little better taste.

Conclusion:

Milk and water both seem to work fine... there's no clear advantage of one over the other.

There's a little bit of debate on how thick you should roll out your biscuits before you make them. To find out what difference it makes, I made two batches:

• Batch 1: Thin (rolled 1/2 inch thick before baking)
• Batch 2: Thick (rolled 1 inch thick before baking)

Baked at 425 F. Four taste-testers.

Taste Testing Results:

Three of four people liked the thick-cut biscuits better.

Conclusion:

Roll 'em thick.

Some people swear by the baking mat, while some swear by the bare cookie sheet. It turns out that for biscuits, it doesn't make a difference.

I made a single batch of six biscuits. I put three biscuits on a baking mat and 3 on a bare cookie sheet and baked at 425 F.

Can you tell a difference? I couldn't... not in taste, texture, or appearance, even on the bottom of the biscuit.

Conclusion:

Don't worry about using a baking mat.

Many biscuit recipes caution against over-working the dough. I've read recipes that insist you only stir 15x after adding the milk and others that say you should knead only 5-10x. The idea is that too much kneading will make the biscuits tough, hard, and nasty.

Usually I mix just until the ingredients are combined, then knead 5 times. But to see if this is really necessary, I made three batches:

• Batch 1: Kneaded 5x
• Batch 2: Kneaded 15x
• Batch 3: Kneaded 45x

You can definitely tell after 15 kneads or so that the dough is changing. It gets harder to fold, stiffer, more "bready". By the time you get to 45 kneads, you're starting to get tired, and the dough is resisting you. Something is happening with the gluten in there, I'm sure.

But how does that affect the final product?

Not as much as you'd think.

As you can see in the pictures, the dough that was kneaded only 5x didn't rise quite as much as the other biscuits. It kind of rose and collapsed. The dough that was kneaded 15x and 45x was more "structurally sound" (like the biscuits we've seen before with less butter or milk).

As for taste and texture, my wife and I noticed two things:

• Kneading more makes the dough more homogenous. In other words, less crumbly, more spongy... like Pillsbury biscuits from a can.
• Kneading more makes the texture slightly tougher. Emphasis on slightly... like, almost unnoticeably.

Conclusion:

Kneading somewhere around 10x is probably ideal, but I seriously doubt anyone would complain if you kneaded 5x or 45x. You're not going to absolutely ruin the biscuits by "overworking" them.

How about baking temperature? Is it better to bake for a short time at a high temperature, or for a long time at a low temperature?

I made a single batch of biscuits and baked two biscuits at 425 F and two biscuits at 350 F. It took a little longer to finish at 350 F (20 minutes vs 15 minutes), but the resulting biscuits were almost indistinguishable. My wife thought the 350 F biscuits were a little tougher. I thought they tasted the same.

Conclusion:

Bake at a higher temperature... not because it makes a big difference in the final biscuit, but because you get to eat your biscuits sooner.

So how do you make the best, flakiest, most optimal biscuit ever?

Well, that depends on your taste. If your taste is like mine, then (for 8 biscuits), use:

• 2 cups of flour
• 1 cup of liquid (milk, buttermilk, or water)
• 1/2 cup of fat (butter, stick margarine, or shortening)
• 1 tablespoon of baking powder
• 1 teaspoon of salt

Stir the dry ingredients together.

Cut in the fat.

Stir in the liquid.

Knead ~10x.

Roll ~1 inch thick.

Bake ~15 minutes at 425 F.

But maybe you just like your biscuits to be as big as possible. If so, up the baking powder, decrease the liquid and butter content, and knead them a bit more.

Or maybe you just want the richest, crumbliest taste and texture possible. If so, use more fat.

Or maybe you're curious about some tests I haven't done yet, like:

• working the dough even less (COMPLETE: see this post)
• milk vs 7-up (apparently it's a thing)
• cutting the butter in vs. melting it and pouring it in
• types of flour
• sour cream (apparently it's also a thing)
• [Fill in the blank]

If that's the case, then check out my blog, where I've posted more installments of my continued quest for the "perfect" biscuit (as well as info on my novels, short stories, other food experiments, and book reviews).

I hope you found this useful. Thanks for reading!

Note: Metric conversions (thanks to ZaKKoS for writing these up):

• 250g flour
• 230ml liquid (milk, water)
• 110g fat (butter, margarine)
• 10g baking powder
• 5g salt
• Roll ~2.5cm thick.
• Bake ~15 minutes at 220° C