At half past six in the morning a poha stand is already three customers deep, and the dish is being finished to order in about the time it takes to count out the change. That is not fast cooking. Nothing is being cooked. The rice in the pan was cooked days or months earlier, in a mill, and what happens at the stand is the last three minutes of a process that began with a sack of paddy.
This is the most useful thing to know about flattened rice: it is not rice that has been crushed. It is rice that has been cooked and then dried flat, and the cooking is the part that makes the speed possible. Starch does not soften in water until it has been taken past its gelatinisation temperature; once it has, the granules are permanently open and will reabsorb water in minutes rather than tens of minutes. A whole grain of rice takes twenty to forty minutes because the heat has to work its way in. A flake a millimetre thick has already had the heat applied, and only needs the water back.
Five of the six steps happen before you buy it
The mill process is short and completely unsentimental. Paddy is soaked in hot water, or steamed under pressure, until the starch gelatinises and the kernel goes from brittle to leathery. It is then passed between heavy rollers, which flatten it and crack the husk off; dried down to a moisture level that will keep in a tin; and sieved to separate the flakes from the bran and the husk. What comes out is a pale, papery, slightly translucent flake that weighs almost nothing and snaps when you bend it.
Thickness is the only variable the product really has, and it is a design decision. Under a millimetre, the flake is translucent and softens almost on contact — good for frying into chivda, useless in a wet dish. Two to three millimetres gives a flake that survives a rinse, a toss and a reheating, and that is the one Indore uses. The choice is not about the rice; it is about what the cook intends to do with it, which is a strange thing for a commodity to encode.
Why a breakfast and not a dinner
Flattened rice is not a prestigious grain product. It is lighter than rice, so it can be carried; it is dry, so it keeps without fuel; it needs no grinding, no kneading and no soaking overnight. Those three properties are what a person in transit wants, and the dish followed the traffic: it is the food of train journeys and pilgrimages, of the market that opens before the offices, of the shift that starts at seven. In a household it is also the insurance in the tin — a month of breakfasts that cannot spoil while you wait for payday.
And there is a cost side that rarely gets said. Cooking fuel is the expensive part of feeding a family in a city, and a breakfast that needs three minutes of flame instead of thirty is not a small saving; it is the difference between eating before work and not. That is the reason the dish is where it is on the social map: not because it is delicious, though it is, but because it is the cheapest hot meal that fits inside a morning.
The garnish is a texture kit
If the base takes three minutes, the interest has to be assembled rather than cooked, and that is exactly what the garnish set is. Sev for crunch, peanuts for fat and crunch, raw onion for bite, curry leaf and mustard for the aromatic top note, lemon for acid, coriander for green, and in Indore a spoon of jeeravan — a sour, spiced powder — plus pomegranate seeds, which exist in the dish for the same reason sev does: to put something in your mouth that is not soft.
That is a different architecture from a curry, and it is why a cart can serve two hundred people in an hour. Nothing is cooked to order except the tempering; everything else sits in a tray and is added in a fixed order, in fixed amounts, by hand. The recipe is really a sequence and a set of ratios, which is what makes it reproducible at speed by someone who has made it ten thousand times.
Two cities, two answers to the same three minutes
Mumbai fries: onion softened in oil, turmeric, the hydrated flakes tossed through, lemon at the end, and the flake is thin enough to go almost pillowy. Indore steams: the hydrated flakes go into a covered pan with the tempering and are heated through rather than fried, and the flake is thick enough to stay separate and chewy. The result, in both cases, is a dish that is recognisably the same and texturally not the same at all — and the difference is one decision, made at the mill, about how hard to press.
The usual story is that the Holkar court carried the dish east when it moved from Maharashtra into Malwa in the eighteenth century, and that Indore then made it its own. Whether or not the court had anything to do with it, the geography is suggestive: poha spreads along the wheat-rice boundary, in the belt where both grains are eaten and neither is sacred. It is the food of the frontier between two staple cuisines, which is perhaps why nobody treats it as a betrayal of either.
The thirty-second window
There is one thing this dish cannot do, and it is the reason it stays on the street. The margin between under-hydrated and mushy is about half a minute, and it depends on the flake, the humidity and the hand. A restaurant kitchen that holds a plate for four minutes under a pass light will send out paste. So poha is excellent at home and excellent at a stand, where it goes from pan to hand in seconds, and it is rare on a restaurant menu outside the places that make it their subject. The technique that made it fast also made it unshippable.
Which is the general lesson, and it is worth more than the recipe. Convenience in food almost never means less work; it means the work has been moved somewhere you cannot see it, and usually to someone who is paid less than you. The flake in the tin is the same invention as the stock cube and the par-baked baguette: cook the starch, dry it, sell the time back. What is unusual here is that the result became ordinary rather than industrial — nobody thinks of poha as a processed food, though every property that makes it work is a processing decision. The most traditional-looking thing on the table is the one with the longest supply chain behind it.
Sources and checks
- Bhattacharya, K. R. Parboiling of rice. In Rice: Chemistry and Technology, American Association of Cereal Chemists, 1985
- Juliano, Bienvenido O. Rice in Human Nutrition. FAO Food and Nutrition Paper 26, Rome, 1993
- Achaya, K. T. A Historical Dictionary of Indian Food. Oxford University Press, 1998
- Sen, Colleen Taylor Feasts and Fasts: A History of Food in India. Reaktion Books, 2015
- Chakraverty, A. Postharvest Technology of Cereals, Pulses and Oilseeds. Oxford & IBH, 3rd ed.
- Collingham, Lizzie Curry: A Tale of Cooks and Conquerors. Oxford University Press, 2006
Last checked 2 November 2026 Yellowsea Huang Corrections