Idli and Dosa Batter Science: A Fermentation Guide

It's Bacteria, Not Yeast, Doing the Work

Idli and dosa batter rises because of a spontaneous lactic fermentation carried out by bacteria already living on the surface of the grains — chiefly black gram (urad dal) — not by any added yeast. The classical account, synthesized in a 1992 National Research Council review that draws on Mukherjee's 1965 work, credited Leuconostoc mesenteroides with producing the CO₂ that leavens the batter, while Enterococcus (then classified as Streptococcus faecalis) supplied most of the acid. In that model, both organisms reach populations above 10^9 CFU/g, and the finished batter settles at a pH around 4.5 after volume increases of 47–113% over 12–20 hours at roughly 29–30°C.

A 2019 culture-independent study by Mandhania and colleagues, published in Applied and Environmental Microbiology (PMC6581174), complicates that story using 16S rRNA sequencing rather than plate counts. Lactic acid bacteria still dominate — about 86% of the community — but the genus Weissella is the real workhorse, averaging roughly 48% of operational taxonomic units and peaking near 71% during the 9–12 hour fermentative window. Firmicutes as a whole climb from about 35% of the population in the first six hours to about 87% by the time fermentation is complete. The authors suggest many older isolates identified as Leuconostoc were likely Weissella misclassified under older taxonomy.

The same study settled a long-open question about which ingredient actually seeds the culture: batters made from surface-sterilized grains failed to ferment at all, and adding a pure culture of Weissella confusa alone was enough to restore normal fermentation. That points squarely at black gram's surface microbiota as the active starter, with rice contributing little on its own. Yeasts have been isolated from batter historically, and Canada's BCCDC still describes idli/dosa starters as a mix of spontaneous yeast and LAB, but the 2019 metagenomic work found their functional role poorly defined compared to the bacteria.

The Official Process: Ratios, Timing, and Temperature

British Columbia's Centre for Disease Control (BCCDC) published formal food-safety guidance for dosa and idli batter in December 2024, and it's the most precise public process specification available. The control points:

• Soak rice and black gram separately in potable water; if the soak runs longer than 4 hours, refrigerate at or below 4°C, otherwise cap room-temperature soaking at 4 hours.

• Grind, then add 2.0% salt by weight.

• Ferment 12–14 hours at approximately 30°C.

• Confirm success with a float/bubble test: batter should roughly double in volume, show visible bubbles, and feel light and fluffy.

• Target a final pH of 4.6 or below before packaging for sale.

• Store fermented batter at room temperature for no more than 1 day, or refrigerated at 4°C for up to 5–7 days.

• Cook to an internal temperature of 74°C for at least 15 seconds.

Fermentation is not just a texture step — it's the food-safety control point. BCCDC ties its pH ≤4.6 target directly to inhibiting pathogen growth before the batter is packaged or served.

Lab measurements back up the timeline. Mandhania's group tracked pH falling from 6.3 to 4.5 and bulk density dropping from 1.2 to 0.5 g/cm³ over 24 hours, with most of the volume increase happening by hour 12–18. A separate study by Ghosh and Chattopadhyay in the Journal of Food Science and Technology (2011) recorded density falling from 0.93 to 0.59 g/cm³, pH from 5.9 to 4.2, titratable acidity rising from 0.44% to 0.91%, and volume increasing 1.62- to 3.2-fold across fermentation.

Ratios and Grind Texture: Idli vs. Dosa

Traditional rice-to-black-gram ratios varied enormously by region and grain price, ranging historically from 4:1 to 1:4 according to the NRC review. BCCDC's current guidance narrows that considerably: roughly 3:1 rice to black gram for dosa and 4:1 for idli. Grind texture also diverges — idli rice is ground coarsely, while dosa batter is ground finer and kept looser so it can be poured and spread thin on a griddle.

Ghosh and Chattopadhyay tested 2:1, 3:1, and 4:1 ratios directly and found the 3:1 ratio produced the highest B-vitamin yield after 7–10 hours of fermentation, with riboflavin reaching 0.76 mg, thiamine 0.73 mg, and folic acid 0.75 mg per 100 g of batter. Black gram alone can leaven a batter on its own, thanks to arabinogalactan proteins and globulins that trap gas as foam; rice alone, per Mandhania's sterilization experiment, essentially does not ferment.

What Fermentation Does Beyond Leavening

Fermentation changes the batter's nutrition, not just its texture. The NRC review reports that stachyose and raffinose — the oligosaccharides responsible for flatulence after eating legumes — are completely hydrolyzed during fermentation, and reducing sugars drop from about 3.3 to 0.8 mg/g. One older estimate cited in the same review claims a roughly 60% increase in methionine content, though the review itself flags this figure as conflicting with other historical measurements, so it's best treated as unsettled rather than established.

Mandhania's metagenomic work found phytase and amylase activity present throughout fermentation (phytase reduces phytic acid, which otherwise binds minerals like iron and zinc), along with a positive plate assay for vitamin B12 production and predicted microbial pathways for B-vitamin synthesis consistent with the measured nutrient gains.

Practical Troubleshooting

Most home fermentation failures trace back to two variables the research keeps pointing at: temperature and grain surface microbiota.

• Cold kitchens stall fermentation. BCCDC's 30°C target is well above typical room temperature in temperate climates, which is why cooks in cooler regions often use an oven with just the light on, or a proofing box, to hold batter near that range for the full 12–14 hours.

• Over-washing or over-sterilizing black gram can kill the starter culture. Since Mandhania's team showed that surface-sterilized grains simply don't ferment, treat unwashed, minimally processed black gram as the source of your live culture, not a contamination risk to scrub away.

• A doubled, bubbly, light batter is the visual proxy for the real target: pH ≤4.6. If batter looks risen but smells off or hasn't visibly bubbled within 14 hours in a warm spot, extend fermentation rather than assuming it's done on time alone.

• Refrigerate promptly once fermentation peaks. BCCDC's storage limits — 1 day at room temperature versus 5–7 days refrigerated — exist because the same acid-tolerant bacteria that leaven the batter will keep metabolizing it past the point of good flavor and texture if left unrefrigerated.

The throughline across both the classical microbiology and the 2019 sequencing data is the same: idli and dosa batter is a live, temperature-sensitive bacterial culture, and black gram's grain surface is where that culture lives.