Restaurant Failure Rate 2026: Debunking the 90% Myth (and What Actually Causes Closures)

“Ninety percent of restaurants fail in the first year.”

You’ve heard it from a banker, a landlord, a well-meaning relative, and probably a business school professor. It’s meant as a warning. It’s also invented. No economist, agency, or academic has traced it to a source.

But here’s the part every debunking skips: the real numbers don’t agree with each other either. You’ll find first-year failure quoted at 17%, 26%, 30%, and 60%, each with a confident citation. They can’t all be right.

They’re measuring different things. Understanding that is more useful than memorizing any one figure.

The Best Available Research

The most rigorous study is a longitudinal analysis of roughly 81,000 full-service restaurants using two decades of non-public Bureau of Labor Statistics business data, by Tian Luo and Philip B. Stark at UC Berkeley (arXiv). They found:

• About 17% of independently owned restaurants closed in their first year.
• The comparable figure for other service-providing startups was 19%, so restaurants were slightly less likely to fail than the average new service business.
• The median restaurant lifespan was about 4.5 years, slightly longer than other service startups.

Separate work by H.G. Parsa at Ohio State, tracking restaurants in Columbus, found a first-year rate closer to 26%: higher, but in the same universe, and nowhere near 90%. The Bureau of Labor Statistics’ business survival data for accommodation and food services broadly supports that range.

So the honest headline: opening a restaurant is hard, about as hard as opening any small service business, and roughly four in five make it through year one.

Why Every Number You Read Is Different

“Closure” and “failure” aren’t the same. A restaurant that closes because the owner retired, sold, lost a lease, moved down the block, or changed concepts didn’t fail. It closed. Most datasets can’t tell the difference.

The unit varies. Some studies count locations, some count companies. A chain closing 40 weak stores is 40 closures and zero business failures.

The time window varies. One year, three, five, “eventually.” Sixty percent over ten years and sixty percent over one year describe different worlds.

Sampling varies. Federal data on hundreds of thousands of establishments and a survey of 200 restaurants in one city won’t agree.

A live example landed this summer. RestaurantData reported about 8,171 restaurant location closures across the U.S. and Canada in the first half of 2026, and noted that among closures with a known opening year, 1,677 had opened between 2024 and 2026, about 40% of the classified total.

That number is already circulating as “40% of new restaurants are failing.” The firm said the opposite: the figure doesn’t show a higher failure rate for new restaurants, because you’d have to compare it with how many opened, and its counts are location closures that include lease expirations, relocations, ownership changes, and conversions.

That’s the machinery of the 90% myth, running in real time.

What Closures Looked Like This Year

The same data contradicts another popular belief. Closures split almost evenly: chain-affiliated locations made up 4,253 (52.1%) and independents 3,918 (47.9%). Some of that is chains pruning weak stores on purpose after years of expansion, which is portfolio management showing up as a closure statistic.

The pressure is real, though. The National Restaurant Association’s 2026 report found 42% of operators weren’t profitable in 2025. Nobody should read “the 90% myth is false” as “things are fine.”

What Actually Causes Restaurants to Close

The same causes surface again and again:

1. Undercapitalization. Opening with just enough for a few months and no cushion for a slow season or a dead compressor. The most common reason a promising restaurant doesn’t see year two.
2. Cash flow mismanagement, distinct from running out of startup money. A restaurant can be profitable on paper and still fail because cash isn’t there the week rent, payroll, and three suppliers all come due.
3. Location chosen on instinct instead of foot traffic, dayparts, and local demand.
4. A concept nobody can describe. If a guest can’t explain your restaurant in a sentence, they can’t recommend it either.
5. Staffing instability, which feeds inconsistent service and food, the two things that stop first-timers becoming regulars.

Notice what’s missing: bad luck. Most closures trace to something identifiable, and several to decisions made before opening day.

What the Survivors Do Differently

Operators who make it past year five tend to open with a real cash cushion, choose location on data, keep the concept narrow enough to execute every night, and spend carefully on everything that isn’t food, people, or rent.

That last habit is where a lot of new restaurants quietly bleed: a pile of monthly subscriptions signed in the excitement of opening, each “only” a few dollars a day. When cash is the thing most likely to close you, every fixed cost you don’t take on is runway.

That’s a reasonable way to look at MenuOnGo if you’re opening or tightening up. Its free plan is a floor, not a countdown: when the 30-day trial ends, nothing is taken down and your menu keeps working for as long as you like. AI builds it from a photo of your printed menu in a couple of minutes, so it doesn’t cost a weekend you don’t have either. If the business grows into needing more, the upgrade is there; until then, it’s one fixed cost you don’t carry.

None of that saves an undercapitalized restaurant. But among the things you control on a Tuesday, protecting your runway is one of the highest-leverage.

Myth vs. Reality

Common Belief	What the Evidence Shows
90% of restaurants fail in year one	About 17% per UC Berkeley/BLS research; 90% has no traceable source
Restaurants are riskier than other small businesses	First-year failure was 17% vs. 19% for comparable service startups
Independents fail far more often than chains	In H1 2026 closure counts, chains slightly outnumbered independents
A closure means a failure	Closures include retirements, sales, lease expirations, relocations, conversions
Failure is mostly bad luck	Most closures trace to capital, cash flow, location, or concept
Surviving year one means you’re safe	Median lifespan is roughly 4.5 years

Frequently Asked Questions

What percentage of restaurants fail in their first year? About 17% of independently owned restaurants, according to the most rigorous research available, a UC Berkeley analysis of roughly 81,000 restaurants using two decades of Bureau of Labor Statistics data. Work at Ohio State put the figure closer to 26%. The widely quoted 90% has no traceable source.

Where did the “90% of restaurants fail” statistic come from? Nobody has found a study behind it. Researchers who have tried to trace it, including hospitality academics, have never located one.

Why do restaurant failure statistics vary so much? They measure different things: locations versus companies, closures versus failures, one year versus ten. Comparing two such figures directly is usually meaningless.

How many restaurants closed in 2026? About 8,171 restaurant locations closed across the U.S. and Canada in the first half of 2026, split roughly evenly between chain-affiliated locations (52.1%) and independents (47.9%). The firm publishing the data cautioned that these are location closures, not business failures.

What is the average lifespan of a restaurant? Research based on two decades of federal data puts the median at around 4.5 years, slightly longer than other service-based small businesses.

---

Sources: Luo, T., & Stark, P. B., “Only the Bad Die Young: Restaurant Mortality in the Western US” (UC Berkeley, using U.S. Bureau of Labor Statistics data), U.S. Bureau of Labor Statistics, Ohio State University research (H.G. Parsa), RestaurantData (via RestaurantNews.com), National Restaurant Association. Images: Pexels (free to use). Data current as of October 2026.