Why this exists.

I kept Googling things like “US cities with population between 50,000 and 150,000” and not finding a clean answer. The results were listicles with no sourcing, scraped Wikipedia tables of unclear vintage, or instructions to download a raw Census file and figure it out yourself. None of them were the thing I actually wanted.

The thing I wanted was simple: pick a population band, get the list, see where each number came from, and download the result as a CSV. That should not be hard. It exists in the official data — the U.S. Census Bureau publishes annual subcounty population estimates — but the path from raw Census file to a usable filtered list is exactly the kind of work most people will not do.

So I built the version that should have existed. The audience is everyone I imagined was running the same search: researchers checking a figure, journalists framing a story, planners sizing a market, real-estate analysts comparing places, hobbyists curious about their state, anyone who wants the answer without trusting a random listicle.

The bar is straightforward. Every number on this site has to be re-derivable from a public Census file. No editorial population estimates, no “approximately,” no synthesized commentary about a city’s character. Just the data, the source, and the methodology.

A normalized, source-traceable register of U.S. cities — the population-side counterpart to a paper Census table you would have asked the reference desk for in 1995. Built on the Population Estimates Program (PEP) annual subcounty estimates plus the Census Gazetteer for geometry.

What’s in

• Current population (Vintage 2024, July 1, 2024 reference date).
• Annual population history: 2010–2019 from V2019 and 2020–2024 from V2024, with the vintage seam clearly marked.
• Within-vintage growth (absolute and percent) — never crossing the seam.
• Land area, water area, density, and centroid coordinates from the Census Gazetteer.
• National and state population, growth, and density rank.
• Per-row source provenance: every value carries a pointer back to the raw file it came from.

What’s not

• Census-Designated Places (other than Urban Honolulu) — PEP doesn’t estimate them annually.
• Minor civil divisions (Edison NJ, Hempstead NY, Brookhaven NY) — a documented V1 gap planned for V2.
• Demographics, race, age, income, housing — those live in the American Community Survey, a separate Census product.
• Metropolitan areas — “New York metro” (~19.5M) is a different geography from “New York city” (~8.5M). City-proper only.

The full list of inclusions and deliberate exclusions is documented on the methodology page. The trustworthiness of this site depends on saying “no” honestly when a use case is outside scope.

I’m Nick. I work in software and data tooling. I run this as a hobby project.

I’m not a demographer. I’m not affiliated with the Census Bureau. I’m a careful operator of a data pipeline that reads public Census files and publishes the result.

The Census Bureau releases a new vintage of the Population Estimates Program each May, revising prior years back to the 2020 census base. When a new vintage ships, this site re-fetches the raw files, re-runs the build, and republishes. The full release cadence is documented at methodology §cadence.

Every City in the USA is independent and is not affiliated with the U.S. Census Bureau. The data is public domain (17 U.S.C. § 105). The site is free to use, requires no signup, and runs no third-party tracking. Census does not review or approve third-party uses of its data, and this site does not imply Census endorsement.