US restaurant universe, mapped
There are 749,404 restaurants in the US. We map each one and the operator who actually buys.
For POS, distributor, payroll, marketing, and equipment teams selling into restaurants. The brand on the sign is rarely the buyer. The franchisee or independent owner in the back office is.
The market, in three numbers
A long tail of independent operators.
are independent single-unit operators
Seven in ten US restaurants are owned by an operator who runs exactly one location. Not a chain. Not a franchisee with twenty units. One owner, one address, one liquor licence.
share held by the top ten chains
Subway, Starbucks, McDonald's, Dunkin', Burger King, Taco Bell, Pizza Hut, Domino's, Wendy's, and Chick-fil-A together account for under a fifth of US restaurants. The other 81 percent is everyone else.
units at the largest brand
Subway runs roughly 20,500 US locations, almost all franchised. Even the biggest brand in the country is less than 3 percent of the universe, and almost none of those sites are corporate-owned.
Sources: National Restaurant Association 2024 State of the Restaurant Industry; Technomic Top 500 Chain Restaurant Report 2024; Orbital site-by-site map, April 2026.
Methodology
Why our restaurant count moves and the published estimates do not.
The headline figure you usually see, 749,404 locations, comes from the NRA's annual State of the Restaurant Industry, which is built on BLS QCEW microdata that lags by five to seven months. It is the right anchor. It is also a snapshot. Restaurants close fast. We refresh the site-level layer between the annual reports.
How the 749,404 figure is built
- Start with every active US eating-and-drinking-place location. Anchored to the NRA's 2024 universe and the BLS Quarterly Census of Employment and Wages NAICS 722 series. Foodservice contractors and lodging foodservice sit outside this number on purpose.
- Resolve each location to a real operating business. A McDonald's on the corner is not McDonald's Corporation. It is usually a franchisee LLC that holds the lease, the liquor licence, and the payroll for that store and several others. The operating company is the buyer.
- Find the operator. Around 70 percent of US restaurants are independent single-unit operators. Most of those owners never built a LinkedIn profile. We find them by name, with a verified email and a direct dial, the same way we do for every other long-tail vertical.
- Drop the dead pins. Closures, rebrands, sites that switched concepts, locations that went dark during the pandemic and never reopened. NRA's annual report holds them on for a year. We do not.
- Refresh on a rolling schedule. Site-level signals run continuously against the universe of US small businesses, so what you query in June is not what shipped in January.
If you want the source breakdown for a specific state, segment, or chain, ask. We do not hide the working.
By state
Where the restaurants actually are.
The five largest states carry around 37 percent of the universe. Per capita, the picture changes: dense Northeastern markets and tourism-heavy states have far more restaurants per resident than the national average.
| # | State | Restaurants | Per 100k residents |
|---|---|---|---|
| 1 | California | 90,400 | 232 |
| 2 | Texas | 58,700 | 189 |
| 3 | New York | 51,300 | 262 |
| 4 | Florida | 49,800 | 216 |
| 5 | Pennsylvania | 31,200 | 240 |
| 6 | Illinois | 28,500 | 228 |
| 7 | Ohio | 26,800 | 226 |
| 8 | Georgia | 24,100 | 216 |
| 9 | North Carolina | 23,400 | 216 |
| 10 | Michigan | 22,700 | 226 |
| 11 | New Jersey | 22,100 | 237 |
| 12 | Virginia | 20,300 | 232 |
| 13 | Massachusetts | 19,400 | 277 |
| 14 | Washington | 18,200 | 232 |
| 15 | Arizona | 17,100 | 230 |
Counts rounded to the nearest hundred for display. The dataset itself is exact, down to the street address. Source: Orbital site-by-site map, April 2026, anchored to BLS QCEW NAICS 722; per-capita math against US Census 2024 population estimates.
The top ten chains
The largest brands, and how little of the market they actually hold.
Brand signs are everywhere. The math is sober. The ten brands below together account for around 19 percent of US restaurants. The remaining 81 percent is independents, regional concepts, and small franchisee groups.
| # | Chain | US units | Notes |
|---|---|---|---|
| 1 | Subway | ~20,500 | Almost entirely franchised. Most US franchisees own one to three stores. The operator is the buyer for almost everything except the national supply contract. |
| 2 | Starbucks | ~16,800 | Predominantly company-operated, with a smaller licensed footprint in airports, grocery, and hotels. The corporate model means a single procurement team for most vendor categories. |
| 3 | McDonald's | ~13,500 | Roughly 95 percent franchised. Franchisees average 3 to 4 stores each, with a long tail of larger operators. The franchisee LLC is the buyer for most local services. |
| 4 | Dunkin' | ~9,400 | 100 percent franchised in the US, concentrated in the Northeast. Franchisees often own 10 to 50 units inside a regional cluster. |
| 5 | Burger King | ~7,000 | Owned by Restaurant Brands International. Almost entirely franchised, with a few large multi-unit operators running hundreds of stores each. |
| 6 | Taco Bell | ~7,200 | Yum Brands portfolio. Predominantly franchised, with mega-franchisees like Pacific Bells running 200-plus units. |
| 7 | Pizza Hut | ~6,500 | Yum Brands. Mostly franchised. Footprint has shifted toward delivery-and-carryout units and away from the dine-in red-roof format. |
| 8 | Domino's | ~6,800 | Franchised at around 99 percent in the US. Franchisees grow up inside the system, so multi-unit ownership is the norm. |
| 9 | Wendy's | ~6,000 | Roughly 95 percent franchised. The franchisee base is concentrated in a handful of large operators across the South and Midwest. |
| 10 | Chick-fil-A | ~3,100 | Single-operator model: every US site has one named operator who runs that one store. Lower unit count, much higher per-unit revenue than peers. |
Counts marked "~" are approximate, drawn from each brand's most recent franchise disclosure document and cross-referenced with our site map. For franchised brands, the buyer for your software or service is almost always the franchisee LLC, not the corporate parent. Source: Technomic Top 500 2024, brand FDDs, Orbital April 2026.
Our take
The chain story is mostly a sign on a building.
We believe
If you sell into restaurants and you only target the brand HQs, you are walking past 600,000 buyers.
The standard vendor motion in this market is to chase the brand. Land Subway corporate, land McDonald's procurement, win the national contract. We watched a restaurant POS team work that motion for three years, then look at the math: Subway runs around 20,500 units, McDonald's around 13,500, and even rolled together with the rest of the top ten, that is roughly 19 percent of US restaurants. The other 81 percent, somewhere over 600,000 locations, sits with independent operators and franchisee groups who never came up in their CRM.
One of those operators called us last quarter. She owns four McDonald's franchises outside Atlanta. Her franchisee group had not been contacted by their POS vendor, their food distributor's tech team, or their payroll provider in over a year. Every meeting her ops manager booked, they had to find themselves. That is the market most vendors are missing, not because the operators are hiding, but because the data tools are searching for "McDonald's" instead of for the LLC that pays the lease.
Who buys this data
B2B vendors selling into 749,404 locations.
This page is for the teams selling into restaurants, not the operators themselves. The buyer for this dataset usually falls into one of these categories.
Restaurant POS, payments, and online ordering
Toast, Square for Restaurants, Olo, and the next wave of cloud POS and ordering platforms. The buyer is the independent operator or the franchisee group, not the brand.
Broadline food distributors
Sysco, US Foods, Performance Food Group, and the regional broadliners. Account-acquisition teams sourcing the next thousand independent restaurants on their delivery routes.
Restaurant payroll, HR, and back-office
Gusto, Restaurant365, Crunchtime, and the rest. The buyer is the operator drowning in 14-day pay cycles, tipped-employee compliance, and food-cost reconciliations.
Restaurant marketing and reputation
Yelp Ads, Bentobox, Owner.com, and the local-SEO and CRM stacks. Operators who run one to five locations are the under-served buyer here, not the chains.
Equipment, supply, and smallwares
Equipment dealers, used-equipment auction platforms, smallwares and chemical-supply route reps. The buyer is the operator opening a second location or replacing a 12-year-old walk-in.
Restaurant staffing and scheduling
Snagajob, Workstream, 7shifts, and the hiring platforms specific to hourly foodservice work. The operator is hiring twelve times a year per location whether they want to or not.
Third-party delivery and reservations
DoorDash, Uber Eats, Toast Tables, OpenTable. The buyer is the operator deciding whether to absorb a 25 percent commission for another year or build first-party ordering.
Adjacent universes built the same way: the restaurant owner email list, the US gas station universe, and the broader by-industry email lists.
Plain-spoken
When the restaurant dataset is the wrong fit.
Do not buy this if any of the following are true.
You only sell into the national brands. If your motion is one annual contract with Yum Brands corporate and one with McDonald's procurement, you do not need 749,404 records. You need a procurement contact list. Save your budget.
You need real-time food cost or commodity pricing. USDA, Urner Barry, and the broadline distributors do that. We map the operators and the locations, not the daily protein spread by region.
Your product is for diners. Restaurant discovery, reservations as a consumer, food-blogger directories: the data here is operator-side, not diner-side. Different shape, different licence.
You need menu or recipe-level data. Menu intelligence sits with Datassential, Technomic, and a few specialist menu-scraping vendors. We tell you who runs the restaurant, not what is on tonight's special.
The honest version
Why most restaurant vendor data is wrong.
If you Google "how many restaurants in the US," the top results split into two camps. The NRA cites 749,404 eating-and-drinking-place locations for 2024, anchored to BLS QCEW. IBISWorld and Statista quote numbers above one million by rolling in foodservice contractors, lodging foodservice, food trucks, and caterers. Both are technically correct, for different definitions. For vendors selling into standalone restaurant operators, the NRA figure is the one that matches the buyer universe. We use 749,404 because that is what a POS deal or a distributor account actually looks like on the ground.
The next problem is the brand. Enterprise data tools index by company, so "McDonald's" looks like one customer with 13,500 locations. It is not. McDonald's corporate owns roughly 5 percent of US stores. The rest are run by franchisee LLCs, often with names like "Atlanta Golden Arches Group" or "Sunshine Restaurants of Florida," and each is a different buyer with a different signing authority on a different contract. The big database returns one row. The reality is several thousand operating companies.
This is exactly the gap Orbital was built for. We map the universe of US small and mid-market businesses, find the operator of each one, and validate the contact before it reaches you. Nothing about that is restaurant-specific, which is why we can also map dentists, HVAC contractors, gas stations, med spas, and auto dealers the same way. What is specific to restaurants is the layer on top: concept and segment, service format, chain or franchisee affiliation, liquor licence, and whether the operator runs one location or fifty.
One more piece of context worth pricing in. National Restaurant Association, BLS QCEW, IBISWorld, and ScrapeHero location reports all publish counts on this market. They are useful, and we cite them. They are also annual, or quarterly with a five-to-seven-month lag. For a vendor doing outbound this quarter, the question is which locations are open this Monday and which operator is on the phone. That is the gap a location-by-location, operator-by-operator map closes. The independents are where the deal flow actually is.
Questions
Before you ask sales about the restaurant dataset.
How many restaurants are there in the US?
There are 749,404 active restaurants in the United States, based on the National Restaurant Association's 2024 State of the Restaurant Industry count of eating-and-drinking-place locations, cross-checked against Orbital's site-by-site map. The figure covers full-service restaurants, limited-service restaurants, snack and nonalcoholic beverage bars, cafeterias, and bars and taverns. Foodservice contractors and lodging foodservice are tracked separately.
Who owns the most restaurants in the US?
Subway is the largest US restaurant brand by unit count at roughly 20,500 locations, followed by Starbucks at around 16,800 and McDonald's at about 13,500. Almost every one of those sites is franchised. The franchisee is the buyer for almost everything except the national supply contract. Even rolled together, the top ten brands account for under 20 percent of US restaurants. The market is structurally a long tail of independents.
Are most restaurants owned by big chains?
No. Around 70 percent of US restaurants are independent, single-unit operators, per the National Restaurant Association. The franchised sites under McDonald's, Subway, or Taco Bell are owned by franchisees who run between one and a hundred locations each. Even at chain-heavy brands, the operating company is usually a separately incorporated franchisee, not the corporate parent. For most vendors, the decision-maker is the operator, not the brand.
What counts as a restaurant in this number?
The 749,404 figure covers eating-and-drinking-place locations as defined by NAICS 722 and reported by the NRA: full-service restaurants, limited-service restaurants, cafeterias, snack and nonalcoholic beverage bars, and drinking places that serve food. Foodservice contractors (think Aramark in a corporate cafeteria) and lodging foodservice (the hotel restaurant) are reported separately in NRA's broader 1,000,000-plus location count. When most vendors say restaurant, they mean a standalone operator. That is the 749,404 figure.
Which state has the most restaurants?
California has the most, with around 90,400 locations, followed by Texas at roughly 58,700, then New York at 51,300, Florida at 49,800, and Pennsylvania at 31,200. Per capita, New York, Vermont, and Hawaii lead. The Sun Belt has been growing fastest, with Florida, Texas, and Tennessee all adding locations since 2020 while the Northeast has been flat to declining.
How accurate is the US restaurant count?
It depends on what you count and when. The NRA State of the Restaurant Industry ships annually and is built off BLS Quarterly Census of Employment and Wages microdata. That is the gold standard, and we anchor to it. The problem is the lag: BLS QCEW data is typically published five to seven months after the quarter ends. Restaurants close fast. Orbital refreshes the site-level layer monthly against the universe of US small businesses, so closures and new builds get caught between the annual reports.
Who actually buys restaurant data?
Vendors selling into restaurant operators. Restaurant POS and payment platforms like Toast, Square for Restaurants, and Olo. Food distributors like Sysco, US Foods, and Performance Food Group. Restaurant-specific payroll and back-office software like Restaurant365 and Crunchtime. Marketing and reputation tools like Yelp Ads, Bentobox, and Owner.com. Equipment dealers, third-party delivery, staffing platforms, and reservation systems all sit on this list too. The common thread is they need the operator, not the brand on the sign.
Can I get a sample of the restaurant owner data?
Yes. Tell us the states, segments, or chain affiliations you want and we send a sample of around 100 verified owner records so you can check them against your own pipeline before anything changes hands. There is no charge for the sample.
See the restaurant operator dataset before you pay for it.
Tell us the states, segments, or chain affiliations you want. We send a free sample of around 100 verified operator records you can check against your own pipeline, no commitment, no email-list back-and-forth.
Get the sample