Methodology
This page sets out, in full, how every figure on Miles Between Cities is produced. We publish it because a distance figure without a method is an assertion, and readers planning moves, trips and deliveries deserve better than assertions. Everything described here is applied uniformly: the same sources, the same formula and the same labelling rules govern the shortest hop in the index and the longest cross-country pair.
Coordinates come from the U.S. Census Bureau Gazetteer files, the federal reference that lists every incorporated place and census-designated place in the United States with an internal point — a representative latitude and longitude — for each. We record the U.S. Census Bureau Gazetteer name, state and coordinates of every city in our index when the data is retrieved (the current edition was retrieved in October 2026) and we use those coordinates, unchanged, as the endpoints of every calculation involving that city. Fixing the endpoints to a public source makes every straight-line figure on the site independently reproducible by any reader with the same file.
Straight-line distance is computed with the haversine formula, which gives the great-circle distance — the shortest path over the Earth's curved surface — between two latitude-and-longitude points, using a mean Earth radius of 3,958.8 miles. Results are rounded to the nearest whole mile. The haversine figure does not depend on roads, traffic or any commercial routing preference, which is why we treat it as the site's anchor measurement: it changes only when the Census Bureau revises a coordinate, and it means the same thing on every page.
Driving distance, where a page or table row is labelled measured, comes from OSRM — the Open Source Routing Machine — run against OpenStreetMap road data. OSRM finds the fastest routable road path between the two Census coordinates and returns the path's length and a duration estimate; where the response names the principal roads, that route summary is published with the figure. Driving distance, where labelled estimate, is calculated as the straight-line distance multiplied by a circuity factor of 1.25, and its drive time as that estimated road distance divided by an assumed average speed of 50 miles per hour. The factor reflects typical American intercity circuity, which generally falls between about 1.1 and 1.4; it is an allowance, not a measurement, and pages never present it as anything else.
Drive time is labelled an estimate on every page without exception, including pages whose distance was measured, because no static calculation includes the traffic, weather, roadworks and stops of a real journey. OSRM durations reflect the routing engine's speed model for the road data; estimated durations reflect our stated 50 mph assumption. Either way, readers should add stops and a margin, as the guides on this site advise, and check live conditions on the day of travel.
Some destinations have no road connection to some origins — island cities and roadless places among them. For those pairs we publish the straight-line distance only and mark the road columns accordingly. We never synthesise a driving figure for a journey that cannot be driven, and we never fill a gap in the data with a plausible-looking invented number: a missing figure is labelled missing, an estimated figure is labelled estimated, and a measured figure names its engine and its data source.
When source data is refreshed, pages are regenerated from the data so that headlines, tables, hub listings and question sections always agree with one another. If you spot a disagreement between two pages on this site, or between our figure and a source you trust, please report it through the contact page with the pair and the figures involved; verified corrections are incorporated into the next regeneration.
Sources: Coordinates: U.S. Census Bureau Gazetteer files; driving data: OpenStreetMap/OSRM — retrieved October 2026.