Kati Roll: How Kolkata Street Food Became a New York Late-Night Staple

What the Roll Actually Is

Before the American version, before the lines outside **The Kati Roll Company** on West 46th Street, there was a pavement counter in Kolkata and a problem of logistics. The original kati roll is credited to **Nizam's**, a restaurant in the New Market area of central Kolkata that has been operating since 1932. The original construction: a seekh kebab, skewered on iron rods over a charcoal grill, pulled off the skewer and wrapped inside a paratha that had been cooked on a flat griddle, brushed with egg, and folded around the meat with raw onion, green chili, and a squeeze of lime. The wrapper was not an afterthought. It was the engineering solution to a city where office workers and mill laborers needed a full meal that required no utensils and left no trail.

The egg-washed paratha is the structural signature. The egg binds the bread on one side, creates a thin layer that acts as a moisture barrier between the filling and the outer layer, and gives the whole package enough grip to be held with two fingers. It is a practical detail that turned out to also be a flavor detail: the egg picks up char from the tawa and adds a faint savoriness that plain paratha does not have. Every New York shop that does this correctly is doing something that was solved in Bengal before World War II.

The fillings evolved: chicken tikka, paneer, egg alone, potato with cumin. The Kolkata street version is not precious about this — it is a format, not a recipe, and the format accommodates most proteins and most vegetables if you cook them right. What it does not accommodate is a filling that is too wet, a paratha that is too thick, or a roll that sits for more than four minutes before it is eaten. New York learned the third rule slowly.

West 46th Street, 2002: The First Counter

**The Kati Roll Company** opened in 2002 in a narrow storefront on West 46th Street, in the corridor that midtown Manhattan calls Restaurant Row. The founder, Payal Saha, had grown up eating rolls in Kolkata and could not find a version in New York that matched the one she remembered. The shop was tiny. There was no dine-in seating in the early configuration, just a counter and a line that, within a few years, was wrapping down the block at lunch and again after midnight when the theater crowd and the late-shift workers came through.

The timing mattered. In 2002, Indian food in Manhattan was concentrated in two registers: the white-tablecloth curry houses of East 28th Street, where the dishes were long-cooked and the portions were designed for groups, and the cheap lunch buffets that occupied the same blocks at midday. There was almost nothing in the fast, portable, under-ten-dollar category that was also Indian. The kati roll occupied that gap entirely. You could get a chicken roll for seven dollars, eat it on the sidewalk, and be back at your desk in twelve minutes. The office workers of midtown noticed this immediately.

**The Kati Roll Company** did not franchise in the early years. It opened a second location in Greenwich Village, then a third back in midtown, and it operated with a degree of consistency that made it a known quantity rather than a novelty. The Kolkata model — fast, cheap, portable, reliable — translated to New York because New York's street-food economy runs on exactly those variables. The dish did not have to be explained to the city. The city already understood what it needed.

By 2010, the model had spread. Not through direct competition with The Kati Roll Company, but through the natural diffusion of a format that other Indian restaurateurs could see was working. Shops in Jackson Heights, in Astoria, in the East Village, started adding kati rolls to their menus. Some called them frankie rolls, following the Bombay variant. Some kept the Kolkata construction. The distinctions between the two versions became a small, ongoing argument among people who cared about the difference.

The Frankie Problem: Two Cities, One Wrapper

The kati roll and the Bombay frankie are not the same thing, and New York has mostly decided not to care about that. This is understandable from a consumer perspective and mildly aggravating from a culinary one. The two wraps share a format — a flatbread, a filling, a wrapper — but the origins, the bread, and the flavor logic are distinct.

The Kolkata roll starts with a paratha: layered, laminated dough, cooked on a tawa with ghee or oil, finished with egg on one side. The Bombay frankie traditionally starts with a maida-based roti that is thinner and less flaky, coated with egg and cooked similarly, but the texture is different in a way that changes how the wrap holds together and how it releases flavor as you eat it. The Kolkata version is richer by construction. The Bombay version is lighter. Both are correct in their context.

New York's Indian restaurant community arrived from both cities, and from Gujarat, and from Tamil Nadu, and from a dozen other places, and the shop that sells "kati rolls" on a given block may be run by someone whose family never ate a kati roll in Kolkata. **Sukhadia's** in Edison, **Maharaja Sweets** in Jackson Heights, and counters inside Indian grocery stores across the five boroughs each have their own version of the wrapper, and the differences compound when you factor in halal certification, vegetarian-only kitchens, and the American tendency to add a chile sauce that has no precedent in either city.

This is not a failure of authenticity. It is what happens when a format is useful and the population that brought it here is internally diverse. The kati roll in New York is not one dish. It is a format held loosely by a common set of techniques, and the variation inside that format is a record of where Indian food in America actually came from. See also the work The Dish explored in our chaat counter coverage, where the same pattern appears with a different dish: a street-food format landing in America and immediately fracturing into regional sub-versions, each one accurate to somebody's memory.

The Geography: Jackson Heights, Jersey City, and the Late-Night Index

The kati roll economy in New York is not centered in Manhattan. It is centered in Jackson Heights, Queens, and in stretches of Jersey City and Newark where the South Asian diaspora built its food infrastructure starting in the 1970s and 1980s. The shops on 74th Street in Jackson Heights, many of them open until 1 a.m. or later, are the most concentrated cluster of the format outside Kolkata itself. **Patel Brothers** anchors the retail end. **Rajbhog Sweets** and **Spicy Mina** anchor the prepared-food end, the latter a counter-service operation with a line that runs from dinner through the taxi-driver rush at 2 a.m.

The late-night function of the kati roll is structural, not accidental. The roll holds its integrity for long enough to survive a commute. It is filling without being heavy. It costs, at most shops in Jackson Heights, between seven and twelve dollars. And the shops that sell it stay open because the neighborhood's working population runs on late shifts, irregular hours, and a food economy that does not close at ten. The roll and the neighborhood's schedule are aligned.

In Jersey City, the same geography applies. The stretch of Newark Avenue from Journal Square toward the Indian restaurant cluster is not midtown Manhattan; it is a working neighborhood where the food is priced for the people who live there, not for the people who commute in from elsewhere. **Roti Roll** and several unnamed counters inside sari shops and sweet stores sell versions of the roll at prices that track the neighborhood's economic reality. This is the same logic documented in The Great American Biryani Belt: the densest South Asian food corridors in America are not the ones the food press writes about most, but they are the ones where the food is most reliable.

The algorithm notices this pattern consistently. The highest-performing kati roll spots in the New York metro area cluster in Jackson Heights, downtown Jersey City, and the Elmhurst-Woodside corridor. The midtown Manhattan shops score well on consistency and speed. The outer-borough and Hudson County shops score higher on flavor and value. The gap is not large, but it is consistent across hundreds of visits, and it points to what happens when a dish is priced for a community rather than for a market.

Why the Roll Works After Midnight

The kati roll is a late-night food by design, not by accident. In Kolkata, **Nizam's** and the dozens of roll counters that followed it were always open past midnight, feeding mill workers finishing the second shift and college students finishing an argument. The physical logic of the roll — sealed, portable, no utensils, one hand — makes it the right food for the hour when restaurants have closed and the city is still moving.

New York's late-night food economy before 2002 was pizza by the slice, halal cart chicken and rice, and diner food. These are not bad options. They are, however, limited in what they do. The kati roll added a category: something hot, spiced with actual complexity, under ten dollars, available from a counter rather than a cart, and eaten standing up in under six minutes. The halal carts deserve credit for proving that South Asian spice profiles could sell to a broad New York audience at midnight. The kati roll proved the same thing at a counter with a fixed address and a menu.

**The Kati Roll Company's** midnight line on West 46th became a documented phenomenon within a few years of opening. Food writers noted it. The explanation is simple: theater workers finishing at eleven-fifteen, bartenders off at two, hospital staff rotating off night shifts. All of them wanted something that was not pizza, and the roll was there.

The mechanics of the late-night roll are specific. The paratha cannot be made ahead and held. It loses its structure after about four minutes. Every shop that does this right is cooking the paratha to order, which means a line that moves at the speed of a tawa, roughly one roll every ninety seconds per cook. A shop with two cooks on at midnight can handle about forty rolls an hour. That is the operational ceiling, and the shops that survive at that hour are the ones that understood the constraint from the beginning and staffed accordingly.

The Economics Work Like This

A kati roll at **The Kati Roll Company** in 2024 costs between nine and fourteen dollars depending on the filling. A chicken tikka roll at a Jackson Heights counter costs seven to nine. The cost of goods in both cases is roughly similar: a paratha requires flour, oil, and eggs; the fillings are either marinated chicken cooked on a grill or paneer cooked on a tawa; the wrapping is done by hand in under a minute. The labor is skilled but not slow. The margin on a well-run roll counter is higher than on most Indian restaurant dishes, because there is no table, no server, no bussed plates, and no food that sits under a heat lamp waiting to be called.

This is the same economic structure as the dosa counter, analyzed in the piece on the all-day labor behind a $12 lentil crepe: a dish that looks simple from the outside has a labor structure that is actually intensive, but that intensity is concentrated in a way that allows a small operation to run efficiently. The dosa batter requires overnight fermentation and a trained hand on the griddle. The kati roll paratha requires a trained hand on the tawa and a cook who knows when to flip. Neither dish is fast food in the American sense. Both dishes price like fast food in the Indian street-food sense.

The gap between the two price points — seven dollars in Jackson Heights, fourteen dollars in midtown — is not a quality gap. The scoring data shows that several Jackson Heights spots perform in the high eighties and low nineties on flavor. The midtown spots score similarly, sometimes slightly lower on value because the price point is higher against a comparable output. What the midtown price buys is location and hours and a certain reliability of environment. What the Jackson Heights price buys is the dish at its least compromised, priced for the people who eat it most often.

The shops that have closed over the past decade did not close because the roll stopped working as a food. They closed because the rent economics of midtown Manhattan and the East Village punish any operation that cannot turn a table. A roll counter that moves forty units an hour at ten dollars a unit generates four hundred dollars an hour in gross revenue. Against a midtown Manhattan rent, that is a narrow margin. The shops that survive are the ones in neighborhoods where the rent is lower, the customer is a regular, and the volume does not need to spike to cover the overhead.

What the Roll Becomes Next

The kati roll has not been gentrified in the way that ramen was gentrified, or the way that the banh mi briefly became a subject of chef-driven reinvention with unexpected proteins and house-made condiments priced at eighteen dollars. There are upscale Indian restaurants in New York that put a version of the roll on the menu, but the format resists elevation in the way that the banh mi does not, because the structure of the roll is tied to its price. A roll that costs eighteen dollars is not a kati roll. It is a product that borrows the kati roll's format and attaches it to a different economic expectation.

The more interesting development is the spread of the format into non-Indian contexts. Food halls in Brooklyn and lower Manhattan now have counters selling "Indian wraps" that use the kati roll's egg-washed paratha as their base but fill it with ingredients that have no precedent in Kolkata or Bombay. This is not automatically wrong. A format is a format. But the distance between those products and the original is large enough that they are not the same thing, and calling them kati rolls misrepresents the source.

**The Kati Roll Company** is still on West 46th Street. **Spicy Mina** is still open past 2 a.m. in Jackson Heights. **Roti Roll** is still on Newark Avenue. The roll that **Nizam's** invented in the 1930s to feed people who had no time to sit down is still feeding people who have no time to sit down, in a city that has more of those people per square mile than almost anywhere on earth. The format survived the commute from Kolkata to New York because it was designed for exactly this kind of city. Place. Time. Price. The three variables that determine whether a street food survives a migration are also the three variables the kati roll was engineered around from the beginning.