Gluten Free Cookie Dough Factory

Wheat flour is awesome, but predictable. I realize that’s part of its charm, but what happens when we swap it out for something else? Like starches, nut flours, and bean powders? I tested a variety of alternative flours, all of them gluten-free, against a control (all-purpose wheat flour) in chocolate chip cookies. I realize that often gluten-free baked goods work best with a blend of different flours, but I wanted to highlight the different properties of each flour, compared gram for gram against wheat flour. I started with the control recipe below, then substituted in the alternative flours for each test cookie. 

TL;DR? (Spoiler Alert) Click here to skip to the end and see how I ranked each flour, from best to worst.

Ingredients:

112 g butter, unsalted, softened
78 g white granulated sugar
78 g dark brown sugar
1/2 tsp vanilla extract
2 eggs, large
144 g flour (various types listed below)
1/2 tsp baking soda
1/2 tsp salt
150 g semisweet chocolate chips

Directions:

Preheat oven to 375°F and line sheet pan with parchment paper. In a stand mixer, cream together the softened butter, white granulated sugar and dark brown sugar on medium speed until light in color and texture (like frosting). Meanwhile, combine the flour, baking soda and salt in a medium bowl and set aside. Once creaming is complete, add the eggs and vanilla extract and beat on medium speed until incorporated. Slowly beat in flour mixture a few spoonfuls at a time. Scrape down sides and mix in chocolate chips on low speed.

Use a 1.5 Tbsp (#40) cookie scoop to deposit cookie dough on an ungreased parchment-lined baking sheet. Dough should be placed 2 inches apart to allow room for cookies to spread. Bake for 9 minutes, or until edges are golden brown. Rotate the pan halfway through cooking. Allow to cool before attempting to remove the cookies from parchment paper.

 Raw cookies, gluten-free experiment

The raw cookie dough gives us some indication of what’s to come. For instance, the two starch-based flours, potato and tapioca are oozy and separating, and coconut looks dry as a bone compared to the control. We can predict some trouble there, but we won’t know for sure what will happen until we bake them.

Baked cookies, gluten-free experiment

The three doughs we singled out as worrisome–potato, tapioca, and coconut–did indeed have trouble. The two starch-based flours, potato and tapioca, spread out like a liquidy mess, whereas the coconut flour didn’t spread or puff at all. In fact, its ridges merely toasted. These three aren’t winning any beauty contests, but the good news is, the rest of the flours resulted in proper looking cookies! Here’s the run down on each cookie.


Control, bakedControl (all-purpose flour)

Appearance: Normal, slightly cakey chocolate chip cookie. Surface is matte with no sheen.
Taste: Flavor is pleasantly sweet and toasty with hints of caramel.
Texture: Soft in texture, but firm enough to hold together.
Would I serve it? Yes.


White rice flour, bakedWhite rice flour

Appearance: Looks the most like the control cookie. Surface has a slight sheen.
Taste: Pleasant sweet taste with fewer caramel notes than the control.
Texture: A slight skin on the surface of the cookie gives it a hint of extra texture. There is a slight grittiness compared to the texture of the control, but nothing too significant.
Would I serve it? Yes.


Brown rice flour, bakedBrown rice flour

Appearance: Somewhat darker in color than the control (looks similar to the almond flour cookie). Surface has a slight sheen.
Taste: Similar to the white rice cookie, but with a few more caramel notes. Still somewhat milder than the control.
Texture: Much like the white rice cookie, there’s a slight skin on the surface which gives it a hint of extra texture. It’s slightly grittier than the white rice cookie, but not by much. Still not too distracting compared to the control.
Would I serve it? Yes.


Almond flour, bakedAlmond flour

Appearance: Somewhat darker in color than the control (looks similar to the brown rice cookie). Surface has a slight sheen.
Taste: The flavor of this cookie is light and sweet with a slight almond character–very tasty.
Texture: Much like the two rice cookies, there’s a light skin on its surface which gives it a hint of extra texture. It’s the grittiest of the alternative flours in this experiment (I used Bob’s Red Mill brand almond flour).
Would I serve it? No, unless I couldn’t use grains. It has great flavor, but the grittiness is distracting, and many of the grain flours make a better cookie. Almond flour would, however, be my first choice among these for someone on a grain-free diet.


Oat flour, bakedOat flour

Appearance: This cookie had a lovely appearance. Although slightly darker than the control, it looks very similar in every other respect. Surface is matte with no sheen.
Taste: The oat flavor comes through in this cookie, but it tastes delicious and appropriate in the context. It’s a lot like a chocolate chip-oatmeal cookie hybrid (yum).
Texture: There is no grittiness to the texture; if anything, it’s a little softer or crumblier than the control.
Would I serve it? Yes.


Garbanzo bean flour, bakedGarbanzo bean (chickpea) flour

Appearance: This cookie looks a lot like the control, with a slight yellow glow that’s only slightly darker than the control.
Taste: Leaves you with a strong bean-y aftertaste that is out of place in cookies.
Texture: The texture is quite similar to the control.
Would I serve it? No. I’ve fed this cookie to people who don’t notice the aftertaste, but I can’t stand it.


Sorghum flour, bakedSweet white sorghum flour

Appearance: This cookie looks very similar to the oat cookie–like the control, but slightly darker. Surface is matte with no sheen.
Taste: Its flavor is slightly different than the control, but it is subtle. More subtle than the almond, oat, and garbanzo flours.
Texture: The texture is soft and free of grit, and it holds together well and doesn’t crumble apart.
Would I serve it? Yes. This is my favorite gluten free cookie of the lot. It’s very similar to the control and doesn’t sacrifice anything on taste or texture.


Potato flour, bakedPotato flour

Appearance: Very thin and spread out in all directions. Chocolate chips are left behind in a clump. Surface has a slight sheen.
Taste: Although there’s nothing offensive about the taste, the texture and appearance are such a failure that the taste is irrelevant.
Texture: The cookie is completely thin with no puff or softness, and is crispy, chewy, and unpleasant.
Would I serve it? No.


Tapioca flour, bakedTapioca flour

Appearance: Much like the potato flour, this cookie is very thin and spread out in all directions. Chocolate chips are left behind in a clump. Surface has a slight sheen.
Taste: Although there’s nothing offensive about the taste, the texture and appearance are such a failure that the taste is irrelevant.
Texture: The cookie is completely thin with no puff or softness, and is crispy, chewy, and unpleasant.
Would I serve it? No.


Coconut flour, bakedCoconut flour

Appearance: Basically looks identical to the dough we scooped onto the parchment before baking. The ridges on the surface toasted, but there was no puffing or spreading at all. Surface is matte with no sheen.
Taste: The texture and appearance are so terrible that the taste is made irrelevant. Despite having the same amount of sugar as the rest, this cookie seems barely sweet at all.
Texture: Somehow this is worse than the two starch flours, but in the opposite extreme. It’s a lot like what I imagine kinetic sand would taste like. It feels bone dry and sandy in the mouth. It’s so drying that I won’t hesitate to classify this cookie as a choking hazard.
Would I serve it? No.


Potato flour + tapioca flour + coconut flour blend

(in equal thirds)

I had the idea to combine the three failed batches, since I had a feeling they might balance each other out. It was a surprising success, and prevented what would have been a heartbreaking waste of cookie dough. If I made these again, I would probably back down even more on the coconut flour, as the cookies dried out quickly during storage and took on the super dry, crumbly character of the coconut flour. After a few days, a glass of milk becomes mandatory with these. Although these have potential, several of the other cookies were better than this one and required fewer ingredients, so I probably won’t bother to fine-tune this recipe.

Appearance: Makes a surprisingly decent cookie! Looks as normal as any of the other cookies. Surface is matte with no sheen.
Taste: Flavor is pleasant but slightly bland. The coconut flour blocks some of the sweetness from coming though. A very subtle coconut flavor is present.
Texture: Soft and pleasant and not too gritty. However, after a few days of storage, these cookies become drier, crumblier, and grittier.
Would I serve it? No, simply because there are several better, easier options. But they’re definitely nothing to be ashamed of if you serve them freshly baked.


 

All the flours, ranked from best to worst:

  • Control (all-purpose flour)
  • Sweet white sorghum flour
  • Oat flour
  • White rice flour
  • Brown rice flour
  • Almond flour
  • Potato + tapioca + coconut flour blend
  • Garbanzo bean flour
  • Potato flour
  • Tapioca flour
  • Coconut flour

Want more cookie experimenting? Take my The Food Lab: Science of Cookies class at CulinAerie in Washington, DC to do your own experiment with chocolate chip cookies and create a custom recipe all your own! (Class is not gluten-free friendly–we focus on conventional flour and other ingredients.)

Tagged with →  

6 Responses to Testing Gluten-Free Flour in Cookies

  1. Robin atterton says:

    I am allergic to grains and am desperate to find out if there is such a cookie recipe combining both potatoe flour and coconut flour that would some how work in making a simple butter or sugar cookie type recipe.

  2. Stephanie says:

    I love your blog and the logical approaches you take! (Just found it today while researching leavening.) I thought it of note to add that I’ve made a moderately successful coconut flour based cookie–BUT it also utilized the addition of almond butter. (I wanted to link to the recipe here, but it feels too much like a promotion of another website, one that is not mine!) Coconut flour simply can’t be subbed one-for-one since it’s so dry in my experience. I’ve tried it in pancake recipes, too, but HATE the coconut flavor that takes over! Just my thoughts on it… :)

  3. Christine says:

    Great tests! Very helpful! :D I bought some sorghum and did similar tests with pancakes – however it had a pretty icky aftertaste. I am not familiar with all these other non-wheat flours, and wonder if it is rancid? Did your sorghum flour have an aftertaste? I keep reading its light and sweet, but no mention of what it tastes like when rancid. Do you have any insight?

    • Amanda says:

      Christine –
      It’s possible your four had gone rancid. Many GF flours are prone to rancidity and off flavors, esp if they are whole grain or nut based. You can keep them fresher longer by storing them in the freezer in an airtight container. It’s also possible that sorghum doesn’t work as well or taste as good in pancakes as it does in cookies. This is a common challenge with GF baking–you may find the perfect alternative flour in one recipe, but it won’t work as well in another recipe. My suggestion is to keep on experimenting and find out which flours work best in your favorite foods, then expand from there. Thanks for reading!
      -Amanda

  4. John says:

    Most enlightening. I’d like to taste a wheat/almond/sorghum combo. And where did you get these different varieties of flour?

    • Amanda says:

      John,
      I found many of them at my local grocery stores (Harris Teeter and Safeway). I’ve seen white rice flour and almond flour in stock just about everywhere, but what else they offer will vary store by store. What I couldn’t find in stores, I bought via Amazon.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.