We mapped 50,000 restaurant health grades.
Here's what we found.

Four cities. Four completely different systems.

Every major city grades restaurants differently. NYC uses letter grades that hang in the window. Chicago just says pass or fail. San Francisco gives a numeric score. London rates you 0 to 5. We analysed them all.

NYC: 24,751 restaurants. Every single grade.

This is every graded restaurant in all five boroughs of New York City. Green dots are A grades. Yellow is B. Red is C. Click any dot to see the restaurant, its score, and what went wrong.

Manhattan has the highest concentration of A grades in this snapshot (82%). The Bronx and Queens are at roughly 76%. The data shows an association by borough; it does not establish why the gap exists.

NYC: The borough gap

Not all boroughs are graded equally. Here's the breakdown of B and C grades — the ones that mean problems were found.

Chicago: The city where 22% fail.

Chicago's inspection system is binary. You either pass or you don't. No letter grades, no partial credit. And one in five restaurants fails.

The red dots are outright failures. The yellow dots are conditional passes — restaurants that passed with conditions that must be corrected. Together, 40% of Chicago restaurants have some form of issue on their latest inspection.

San Francisco: Only 41% score above 90.

SF uses a 0–100 scoring system — the most granular of any city. A perfect score is rare. Most restaurants hover between 80 and 95. Below 80 means serious problems.

London: A high share of top ratings.

70% of Westminster restaurants have the top rating of 5. Only 5% are rated 2 or below. London's system is based on penalty points — and most restaurants keep their score low enough to stay at 5.

Temperature control is a recurring inspection risk.

Temperature control, handwashing, pest control, clean food-contact surfaces and adequate food-safety procedures recur in official inspection rules and records. But the agencies categorise them differently, and the UK ratings feed does not publish a comparable item-level violation table.

Cuisine categories show different grade distributions.

In the NYC snapshot, the share of B and C grades varies across the cuisine categories recorded by the city. Those labels are broad and the result can reflect many confounders, including business size, location, inspection frequency and the way cuisines are classified.

This is not evidence that any cuisine is inherently less safe. Operational complexity is one plausible hypothesis, but this descriptive analysis does not test causation. The defensible conclusion is narrower: food businesses should build controls around their actual menu, process and risk profile rather than a category average.

Methodology, sources and limitations.

The numbers in this article are a preserved April 2026 snapshot. The embedded maps query live agency data and may therefore show different totals today.

Limits: agency records can contain duplicates, delayed updates, missing coordinates and classification changes. Business-name/address deduplication is less reliable than a stable agency identifier. Cuisine-category results are descriptive and not causal. Contact hello@blueroll.app with corrections.