Basic steps for homemade tempeh

Black-bean-and-black-rice tempeh after four days.

I update this post from time to time as I learn more from experience. Most recent update: February 11, 2023, to include links to marinades.

Tempeh is a traditional fermented Indonesian and Malaysian soy food prepared from soaked and cooked soybeans by salt-free aerobic fermentation using the mold Rhizopus oligosporus, a fungus. During the fermentation, the dense cottony mycelium produced by Rhizopus binds the soybeans together to form a compact cake. This page has a selection of excellent and informative articles about tempeh.

Tempeh has excellent nutritional value — much better than tofu, also made from soybeans because tofu is more highly processed and to some degree refined. In terms of nutritional value, edamame (cooked fresh soybeans, a whole food) is better than tofu, and tempeh is better than edamame (because tempeh includes both the beans and the fungus mycelium).

I make Malaysian-style tempeh, with the soybean hulls left on. (Indonesian-style tempeh removes the hulls.) I leave the hulls on because (a) it’s easier, and (b) the hulls have nutritional value. (For an interesting discussion of Indonesian-style tempeh, see this article, which includes some cool photos.)

In my experience, homemade tempeh is much better than packaged tempeh I buy in a store — plus it’s satisfying to grow your own, and you can experiment by using different beans, boiled raw peanuts, cooked intact whole grains, or a mix. 

I started a while back, and I’ve gradually identified the success factors to the point that I can now make a good batch on purpose. Here’s a summary of my lessons learned and the method I now use:

1. I often use soybeans, but you can use instead (or in addition) black beans, kidney beans, pinto beans, and so on. You can mix in cooked intact whole grain (spelt, kamut, brown rice, black rice, or the like) or boiled raw peanuts. Avoid using spices and herbs, many of which have anti-fungal properties (one reason they are used in making foods such as sausage: they act as a preservative).
2. Cook beans without using baking soda or salt — use plain water.  (I had a series of failures until I realized that the baking soda was making the beans alkaline, and Rhizopus requires an acidic environment.) 
   .
   Cooking method: Soak beans overnight, drain soaking water using a sieve, then put the beans in a pot and cover well with water and simmer (not boil) until done. I’ve found that beans are not so apt to boil over if I don’t cover the pot. However, with the pot uncovered, it is usually necessary to add water during cooking because of evaporation. Test beans as you cook, and stop cooking before they become soft.
   .
   Cooking time: The time required to cook the beans will vary by the kind of bean, how old/dry the beans are, the altitude (water boils at a lower temperature at high altitudes), and probably the hardness of the water. Lentils, for example, usually require only around 15-20 minutes; soybeans can take more than an hour. Experience will be the best guide, but before you have the experience, use a wooden spoon to remove a few beans from the cookpot and chew them. (A wooden spoon is better than a metal spoon for this since a metal spoon will be much hotter.) If they are hard to chew, cook longer; if they are reasonably tender, they’re done. Do not cook until they are mushy. I test lentils after 13 minutes, soybeans after 45 minutes. Continue to test from time to time until you find the beans edible.
   .
   When beans are ready: Drain beans in a sieve. Do NOT rinse with cold water to stop cooking: you want to use the heat of the beans to help dry them (another reason you end cooking before beans are soft). I used to make a batch of 2 cups of uncooked beans, but ultimately I decided that a 3-cup batch is the ideal size for the bags I use (large Ziploc Fresh Produce bags, perfectly perforated for making tempeh). I often combine 1.5 cups beans and 1.5 cups intact whole grain, measured before cooking and cooked separately. (I have a 1.5 cup measure for dry ingredients, which is handy.)
3. Rhizopus prospers in an acidic environment, and that is achieved by using vinegar (apple cider, white, brown rice, and wine vinegar all will work: the acidity is what’s important). You can choose between two times to add vinegar:
   .
   (a) toward the end of cooking (for the last 15-20 minutes), in which case add 2 tablespoons of vinegar per cup of uncooked beans (thus 1/4 cup vinegar for a 2-cup batch; 1/4 cup + 2 tablespoons for a 3-cup batch) and continue cooking until beans are done; or

   (b) after the beans have been cooked, drained, dried, and cooled,  in which case use 1 tablespoon per cup of uncooked beans (2 tablespoons for a 2-cup batch, 3 tablespoons for a 3-cup batch).
   .
   I now always use method (b), adding vinegar after beans are drained and dried. If you do add vinegar to the cooking water, don’t add it until the very end since acidic water makes the beans hard (and alkaline water makes beans soften quickly, the reason baking soda is used).

   Update: Note the ratio suggested in 6 below: 1 tablespoon for 1kg (roughly 4 cups beans, measured before cooking). If that’s right, I’ve been using four times as much vinegar as needed. OTOH, the way I’ve been doing it works fine and tastes good. Perhaps  I’ll experiment with using less. /update

4. The best way I’ve found to dry cooked beans is to drain them through a sieve, dump them onto a clean dishtowel while they are still hot, and then use a crumpled paper towel to spread them out, pressing the beans gently and rocking them back and forth. Try for a layer that is 1 bean deep, though for me the beans are usually 2 beans deep even when spread out. (You can also use a hair dryer, but be careful not to get it too close to the beans or the blast of air will blow the beans off the towel.) Rhizopus likes a little moisture but definitely not so much that the beans are wet. (I follow the same process to dry grains, which I cook separately.)
5. Leave the beans on the towel until they’ve cooled. The common advice is to let them cool to 95ºF (35ºC), but it’s simpler just to let them cool until they are close to room temperature. Being cooler is not an issue, since they’ll warm up as they are incubated (in an incubator, oven, or whatever).
6. After the beans have been dried and cooled, transfer them to a bowl. (You put the beans on a dishtowel instead of paper towels because a dishtowel is strong enough that you can lift it by the corners to carry the beans to the bowl.) If you’re adding vinegar using method (b), add 1 tablespoon of vinegar per cup of uncooked beans. Then add either one packet of tempeh culture or the proper amount from a bulk package of starter. Add the culture starter a little at a time, using a silicone spatula to mix beans and starter thoroughly after each addition. Even distribution is important for even initial growth of the mold (though uneven growth will normally even out over the growing period.)
   .
   It’s much less costly to buy a bulk amount of starter culture rather than individual packets of starter. Here are sources. However, the link for Tempeh Lab in that list is broken. (The link here is good.) Another source is TempehStarter.com. — Recently, I received a sample tempeh starter from TopCultures, and I used it in making this batch. It has proved to be vigorous and effective, and when I re-order, I will re-order from TopCultures. TopCultures does provide instructions on making tempeh, but they go to the trouble of removing the hulls of soybeans. That step is totally optional, and if you do remove the hulls, you lose their nutritional value. (With other beans, such as black beans, kidney beans, lentils, etc., the hull-removing step is omitted, so obviously it’s not necessary to remove the hulls. That is simply the custom in Indonesia (though not in Malaysia).)
   .
   NOTE: Very cheap cultures may make bad tempeh — not setting correctly, sporing (turning black) immediately, and so on. In the Tempeh Growers group on Facebook, a person told of their experience after buying a cheap tempeh starter culture on eBay. They ended up throwing out six batches of tempeh because the bad starter produced tempeh that did not set up and that spored (turned black) immediately. Saving money on starter culture proved costly in practice.
   .
   I’ve read a suggested ratio of 1kg (2.2 lbs) soybeans with 2 grams starter and 1 tablespoon vinegar. I’ve been using 1 tablespoon of vinegar for each cup of soybeans (measured before cooking). Two cups is roughly 1 lb, so I’ve been using about 4 times as much as they recommend. I’ll perhaps experiment with reducing the amount I use and see how that works. What I have been doing works, though, and it produces good tempeh. 
7. You can spread the beans in a layer in a flat dish (glass or plastic or silicone), but I generally use a large Ziploc Fresh Produce bag because those bags are nicely perforated for ventilation. Also, if the beans are in a dish and not bagged or wrapped, I found that the mold produces a fuzzy growth that puts some off. If the beans are put in a bag or wrapped, the mold forms a velvety smooth coating like the coating on Camembert cheese instead of a surface fuzz.
   .
   The Ziploc bags are also a good size for a batch of 2, 3, or 4 cups — the greater the quantity, the thicker the resulting slab. Spread the beans evenly through the bag, which lies flat on its side, so that the bag is filled side to side and top to bottom, the layer is a nice thickness throughout. (I found that when I lifted the bag to put it into the incubator, that even layer became very uneven. I now put the bag on a wire-mesh rack and even out on that, then transfer the rack (with the bag in place on it) to the raised rack in the incubator.)
   .
   I like the thickness of the 3-cup batch. I did try a couple of 4-cup batches, but found the slab too thick for easy storage or use. On the whole, I find the 3-cup batch more to my liking. 
   .
   I cut through a section of a thick (3- or 4-cup) slab before cooking it to make two (or three) thinner slabs. The 2-cup slab is a little too thin to do that. 
   .
   I plan to stick with 3-cup batches from now on, but I’m glad I tried other sizes, and it was good that I started with small batches and gradually increased batch size. That enabled me to get experience with smaller batches before trying bigger batches.
8. Incubation can be done initially in an oven on the proofing setting (if your oven has that — the proofing setting is specifically intended to promote the growth of a fungus (yeast)) or in an oven with the light on and the door slightly open. For various reasons, I decided to build my own tempeh incubator using rigid 1″ foam insulation board (sold in 24″ squares for projects; three of those squares provide enough, once cut, to make an excellent incubator — details at the link). As noted above, I lay the bag on a flat rack, get the contents into a layer of even thickness, and then place that on a raised rack over a seedling warming pad (details at the link), whose thermostat lets me regulate the temperature.
9. Incubate at 88ºF (31ºC) until Rhizopus is well started. This will take at least 12 hours and maybe 24. My most recent batch took 18 hours before the mold started to appear. Eventually, you will see white patches of mold that look like steamy spots inside the bag. At that point reduce the thermostat to 77ºF (25ºC) and allow the batch to continue to work for another 24-48 hours. That worked for 2-cup batches, but for larger batches, I had to remove the batch from the incubator altogether and put it on a raised rack on the table because the larger batches generate so much heat that they raise the temperature inside the incubator too high.
   .
   Generally, after a couple of days (48 hours) the mycelium completely covers the batch. At that point I let it continue to grow for another 24 hours so the mycelium will solidify the slab. Completion time varies somewhat, but it’s usually 3 days, occasionally 4, of total incubation time once the batch has been started.
   .
   For the first few batches, I did not reduce the temperature, and as a result, the mold spored, which results in dark grey or black areas. The tempeh is still good and perfectly edible, but some find the black spots off-putting. (See “Can You Eat Tempeh with Black Spots?“) That hasn’t happened since I started reducing the temperature once the mold is established.

I started with 2-cup batches and now routinely make 3-cup batches. I did make some 4-cup batches, but now I plan to stick with 3 cups, and I plan to use 1.5 cups legumes and 1.5 cups grain (both measured before cooking and cooked separately). (I use a 1.5-cup dry measure, which is handy.) The green-lentil tempeh linked below was a little over 3.5 cups before cooking.

I’ve noticed that when the batch is out of the incubator (after the first 24 hours), sometimes gray areas will appear on the bottom where the fungus has started to spore. When I see that, I turn the slab in its Ziploc bag over, and the gray areas are soon overgrown, becoming white from new growth of mycelium. I usually turn the slab about every 8 hours.

When you consider the batch done, remove the block of tempeh from the bag, cut it into pieces that will fit your storage containers, and refrigerate it. 

Here are some recent batches; unless otherwise mentioned, the culture starter is from Cultures for Health:

• Black bean and black rice
• Black bean and whole-grain rye
• Black-eyed pea and peanut
• Soybean and unpolished kodo millet
• Chickpea and peanut
• Green lentils 
• Chickpea and black rice
• Pinto beans and khorasan kernels
• Soybean and rye
• Chana dal and unpolished barnyard millet
• Beluga lentils and foxtail millet
• Soybeans and hulled barley
• Kamut and chana dal
• Soybean and Kamut – best batch yet in terms of fermentation
• Red kidney bean and foxtail and little millets
• Soybean and oat tempeh, with a little rye
• Du Puy Lentils and Karmut® wheat, with TopCultures starter
• Chickpea and rye, with TopCultures
• All soybeans, with TopCultures (Malaysian style: soybeans intact, hulls in place) – failed
• Soybean and rye, again TopCultures and soybeans intact – success; grain must help

I have now added the category “Tempeh,” which I use for posts about making tempeh or that contain a recipe that uses tempeh. You can browse posts in that category with “Search Categories” using the category “Tempeh” in the column at the right.

Regarding temperature

For a 2-cup batch (a batch made from 2 cups of uncooked beans), after initial incubation at 88ªF, I just reduced the thermostat to 77ºF and finish the batch. When I made a 3-cup batch (2 cups uncooked black beans and, cooked separately, 1 cup black rice), after the initial incubation at 88ºF, I found that to keep the temperature low enough, I had to remove the batch from the incubator altogether and put it on the table, on a raised rack, and just leave it at room temperature for the rest of the time — 3 or 4 days is what I usually do. See the discussion of temperature in this post. Update – The  3- cup batch came out fine. I also tried a 4-cup batch; it worked, but I settled on a 3-cup batch as the best size for me, and I now make it using 1.5 cups beans/lentils and 1.5 grain (both measured before cooking; cooked separately).

Pictured below is a 2-cup batch of soybean tempeh in process in my homemade tempeh incubator. The left photo was taken after about 12 hours at 77ºF (25ºC). I let it work for 24 hours more, and the result is shown on the right. Nowadays I usually let the tempeh develop for a total of about 3 to 4 days, from start to finish.

Cooking tempeh

I sometimes cut off a piece and fry it like a hamburger patty in a little olive oil and use it to top a dish. As my batches became larger (from 2-cup to 3-cup to 4-cup), the slab of tempeh became thicker. When I have a thick piece of tempeh, I cut it in half or in thirds to make thinner slabs. After frying a thin of tempeh, I sometimes cut it into squares to use like croutons in a salad. 

You can sprinkle squares of fried tempeh with salt and some garlic powder to add flavor, or marinate thin strips (for example, to make tempeh bacon — marinades in general work well with tempeh). I also make tempeh breakfast sausage. Okonomi Kitchen has 5 good marinades for tempeh,  and these 26 tofu marinades at Vegan Foods and Living work equally well with tempeh.

Most often, I dice a portion of a piece of tempeh, enough for a meal. I use small dice for stir-fry, larger dice (bite-size) for stews, curries, and chilis. The stir-fries need not be complicated — for example, I might cook some tempeh diced small with a few chopped cherry tomatoes and onion chopped fine; generally, though, use more ingredients — asparagus, bell pepper, jalapeños, ginger, garlic, chopped red or green cabbage, etc.

A page by Better Nature has links to a wide variety of tempeh recipes (66 as of today).

I have found, BTW, that spraying the pan with olive oil uses much less oil than pouring in a little oil — see this post for the best sprayer I’ve found: 1/4 teaspoon per spray.) I frequently marinate tempeh (for example ponzu or soy sauce, toasted sesame oil, hot sauce, vinegar or lemon juice, grated ginger, crushed garlic, and sesame seed — but it’s easy to make up your own mix). Tempeh responds well to marinade.

After making a series of stir-fries (you can browse recipes with tempeh to see some of them), I started to want some sort of sauce on them. I found a good collection of sauce advice and recipes. See also these recipes, ordered from oldest to newest.

The difference between eating tempeh and eating mushrooms

At right is a piece cut from an early batch of soybean tempeh — the first batch where I knew exactly what to do and why. (The photo looks a lot like a candy bar, doesn’t it? But those are soybeans, not peanuts.) That is a 2-cup batch containing only soybeans (Malaysian style (with the hulls), not Indonesian style (with hulls removed)).

The short, soft velvety coating is like the covering you see on Camembert, though the fungus on tempeh is Rhizopus oligosporus, and the fungus on Camembert (and Brie) is Penicillium camemberti. In either case, beans or cheese, the fungus colony forms a soft white crust. The job of Rhizopus, however, differs from that of Penicillium: the fluffy, white mycelium of Rhizopus welds the together beans (and grain, if that’s included) to create an edible “cake.”

The fungus is why tempeh is more nutritious than edamame: with tempeh, you consume both the bean and the fungus mycelium. Note the difference between Rhizopus and the fungi we eat as mushrooms. When you eat mushrooms, you eat the spore-bearing bodies of the mycelium and don’t even see the mycelium, the actual fungus of which the mushroom is merely the “fruit” that bears the spores. In the case of mushrooms, the mycelium remains hidden underground — but when you eat tempeh, you eat the mycelium itself.

Rhizopus sporing bodies don’t even appear if the tempeh maker knows what they’re doing (as I now do). They form when the tempeh is incubated at too high a temperature. Though edible, their appearance (black or dark gray) is off-putting to some.  (If your tempeh develops patches of color — not white or black or gray but rather pink or green or the like — it’s spoiled and should not be eaten. I’ve never had that happen.)

To prevent sporing, keep the incubation temperature at around 77ºF (25ºC) once the tempeh fungus colony is established. In the photo of the sample above, the snow-white mycelium is clearly visible.

Update: I published a version of this post in Medium, and that version includes photos of my chickpea-and-peanut tempeh.

Mastodon