Making soy milk — Part 2

I didn’t mean for yesterday’s post to be a teaser.  I really thought I’d actually be writing about making soy milk, but when I sat down to write, there was just too much background and context for one post.  But I’m back and ready to get down to business!

The machine
I borrowed my friend’s Soyabella soy milk maker.  The website has a number of recipes for both soy and nut milks, though you have to go to the product manual to get the weight of soybeans to use for the most basic soy milk recipe.


Making soy milk
At it’s most basic, soy milk requires soy beans and water.  You start with dried soybeans (available in the bulk section at most natural food stores).  I had not one, but two small bags of old dried soy beans sitting around, from other times I’d intended to make soy milk, but I decided I should start with a fresher product. The Soyabella manual called for 70g-95g (2.5oz-3.5oz) dry soybeans per 1L batch.


As with all dried beans, the first step is soaking, either a quick soak, or an overnight (at least 8 hour) soak.  Once I had my hands on the machine, I wanted to make soy milk that day, so I went with the quick soak (add enough water to cover beans by an inch, bring to a boil, simmer 2-3 minutes, then remove from heat and soak 2-3 hours).


After that, you add the soaked, drained beans to the grinding chamber, add the appropriate amount of water to the pitcher (I used filtered), push the “Milk” button, and voilá, about twenty minutes later, you have soy milk.

As suggested, I added a pinch of sea salt.  I also reluctantly added a bit of sugar (I switched to buying unsweetened soy milk about three years ago, and I don’t miss the sugar (7g per cup in my usual brand/variety) at all).  However, there’s a very good chance that there’s something sweet-tasting (stevia?) buried in that “natural flavor” portion of the ingredient list.  For my homemade version, I started with about 1 T. of sugar in 1 quart of soy milk, which comes out to < 3g sugar per cup of milk.


Initial results
The initial results were rather, “eh.”  It tasted very beany.  Granted, it is made from beans, but somehow, the soy-bean-only store-bought versions (TJ’s and WestSoy), while not exactly delicious, lack the pronounced beany flavor in my homemade product.

Inspired by a friend’s very positive review of Califia Farms Coconut Almond Milk, I had purchased some coconut cream in hopes that it might lead to an equally delicious soy milk.  (I did not try the Califia product myself, but I glanced at it in the store, and was duped into believing that it consisted of three ingredients: water, almonds, and coconut cream.  I like to think I’m a pretty good label reader, but a glance at the product link above reveals this milk is not as “pure” as it claims (for shame!):

Contains Less Than 2% Of The Following: Sunflower Lecithin, Sea Salt, Potassium Citrate, Natural Flavors, Gellan Gum, Carrageenan, Vitamin/Mineral Blend (Calcium Carbonate, Vitamin A, Vitamin E, Vitamin D2, Vitamin B12, Vitamin B2), Monk Fruit Concentrate.

So, maybe the coconut cream is not really the secret to this tasting good.)


Anyhow, I bought two different brands of coconut cream, and a coconut cream powder.  In the end, these helped a bit, but they were no miracle fix for the beany milk.  This batch was okay for making oatmeal and using for pancakes and waffles, but not inspiring otherwise.  It was usable on cold cereal, but didn’t pass the sipping test.

Other varieties
Not yet dissuaded in my quest for tasty, homemade soy milk, I started playing around.  My second batch of milk in the Soyabella was actually the walnut and almond milk, which is soy free.  The taste was decent (no beans = no beany taste), but it quickly separated into a particulate-heavy bottom and a watery top.

I did some research into eliminating the beany taste in homemade soy milk, and settled on the following fixes:

  • removing the hulls from the soybeans
  • adding a few peanuts to the blend
  • additional cook time for the finished milk

I first tried a batch just removing the hulls.  It tasted better, but I wasn’t satisfied.  The next round, I cut back the amount of soy beans and added a few [soaked] peanuts.  Once the soy milk maker did it’s thing, I dumped the milk into a pan and simmered on very low for about twenty minutes.

The result was WEIRD.  It might have tasted okay, if I was able to get past the texture, but, no.  I’m guessing the stove top cooking, in addition to the cooking that happens in the maker, overcooked it, because the result was ridiculously thick and gloppy.  Adding more water after the fact didn’t help.  I got through this quart by using it primarily for oatmeal.  It was not good for cold cereal or sipping, and I didn’t want to waste good chocolate by using it for hot chocolate.  (In the interest of full disclosure, my other variable here was that I didn’t use only soy beans — I mixed things up a bit by using some black beans, but I really don’t think that was the problem.)

Another method
Some of my research into reducing the beany taste led me to recipes that don’t require a soy milk maker (here and here).  These recipes, which had seemed so daunting before, didn’t look so bad now.  The basic process was the same, minus the convenience of a single machine: soak the beans, grind with hot water (in a blender), filter (using a nut milk bag), and cook (on the stove top).

In the Soyabella, the milk “cooks” while still in contact with the filtered soy bean remains (known as okara).  I was curious if this arrangement (which you can’t avoid if you’re using the machine) intensified the beany flavor.  Armed with a blender and a nut milk bag for straining, I was ready to ditch the machine.


70 g beans (I used mostly soy beans, but I substituted a few chick peas), soaked and hulled
20-30g cashews (raw, unsalted), soaked
1 L boiling water

I pureed the [soaked] beans, nuts, and boiling water in two batches in the blender; filtered it through the nut milk bag; and simmered it on low, stirring frequently, for about twenty minutes.  I added a pinch of salt and 1 T. of sugar, and, hold on to your hats . . . .

. . . . it tasted good!  I actually enjoyed sipping this milk, and it passed the cold cereal test, wonder of wonders!

Back to the machine
To test things a bit further, I returned to the soy milk maker and used the same bean-cashew blend as above.  The result was okay, but, for better or worse, definitely not as good as the non-machine version.

On the one hand, I don’t have to run out and drop $100+ on a soy milk maker.  On the other hand, while making soy milk without a dedicated machine is not as daunting as I’d feared, the machine version requires less hands-on time than the blend, filter, heat and stir on stove top method.  Adding that time up every week could make the not-quite-as-tasty machine version an okay compromise.

Check back for one more post on final thoughts including ease of clean-up, using the okara, cost comparison, and health effects of going “carrageenan-free.”





  1. Stephen Parquet says:

    Non-Beany Tasting Soymilk Recipe

    The ONLY way I have ever found to NOT get beany tasting soymilk is:
    1. Boil on low – 2 cup of unsoaked soybeans for 1.5 hours to 2 hours.
    2. Remove and rinse – skins on/off, doesn’t matter.
    3. Put half of the cooked soybeans in blender with 5 cups water (or less if you water stronger soymilk) and blend for a few minutes.
    4. Strain through a cut off and tied panty hose leg piece a few times (stretch over a bowl, tied end down). *Even the finest wire mesh is not fine enough.
    5. Repeat steps 3 and 4 with the remaining cooked beans.
    6. That’s it! Makes about 2 quarts.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s