JoseFitani
distance

How to Calculate the Distance Between Two Cities

Distance between cities depends on what you mean by 'distance'. Learn the four methods: great-circle, driving, flight, and rail.

🧭JoseFitani Updated 2026-10-08 6 min read
A great-circle arc compared with a rhumb line between two points.
A great-circle arc compared with a rhumb line between two points. Photo: Jacob Rus (CC BY-SA 4.0), via Wikimedia Commons.

Ask for 'the distance' between two cities and you will get different answers depending on who you ask β€” a pilot, a driver, and a geographer all mean different things. This guide walks through the four main ways to measure intercity distance, explains the math behind each, works a full example, and helps you pick the right method for trip planning, logistics, or analysis.

Method 1: great-circle (as the crow flies)

The great-circle distance is the shortest path along the Earth's surface between two points β€” the distance 'as the crow flies'. It is computed with the Haversine formula, which takes the two latitudes and the longitude difference and returns the central angle between the points, multiplied by the Earth's radius (6,371 km). The formula: a = sinΒ²(Δφ/2) + cos Ο†1 Β· cos Ο†2 Β· sinΒ²(Δλ/2); distance = 2R Β· arcsin(√a). It treats the Earth as a sphere, which introduces at most about 0.3% error versus the more complex ellipsoidal Vincenty formula β€” negligible for almost all purposes. Use great-circle distance for aviation, shipping estimates, radio range, and any rough 'how far apart' question. It is the baseline every other method compares against. For most everyday purposes this spherical approximation is more than adequate.

Method 2: driving distance

Driving distance follows the actual road network and is computed by routing engines (OSRM, Valhalla, Google Maps, HERE) that search millions of road segments for the fastest or shortest path. It is typically 1.2 to 1.5 times the great-circle distance β€” the ratio is called circuity β€” because roads detour around mountains, water, cities and borders. Island and coastal cities have high circuity; cities on flat plains with grid roads have low circuity. Two caveats: routing engines need current map data (new roads change answers), and 'driving distance' depends on the chosen route β€” fastest versus shortest can differ by 10% or more. This is the number you want for road trips, delivery planning, fuel budgeting and moving quotes.

Method 3: flight distance

Aircraft cruise along great-circle routes β€” the shortest path β€” so flight distance starts from the great-circle number. In practice, real flight paths deviate: air-traffic routings follow published airways, winds push aircraft off the ideal track, and climb/descent add distance at both ends. A common planning rule is to add about 5% to the great-circle distance for a realistic flight-path length, more on routes with strong prevailing winds or congested airspace. Flight time then depends on aircraft speed, winds aloft (a jet stream tailwind can cut 30+ minutes off a transatlantic flight), and taxi/hold time. Use flight distance for airfare estimates, carbon calculations and travel-time planning.

Method 4: rail distance

Rail distance follows the fixed rail network, which is often more circuitous than roads β€” railways were laid out in the 19th century around terrain and land ownership, and high-speed lines excepted, trains cannot reroute. Rail distance between two cities is frequently 1.3–1.8Γ— the great-circle distance, sometimes more where the network is sparse or indirect. It matters for train-pass planning, freight logistics and realistic rail journey times: a 500 km great-circle hop can easily be a 700 km rail journey. Unlike driving, you cannot choose an alternative rail route, so the number is what the timetable says it is β€” check the operator's published distance or derive it from scheduled time and average speed.

Worked example: New York to Los Angeles

Take New York (40.71Β° N, 74.01Β° W) to Los Angeles (34.05Β° N, 118.24Β° W). The great-circle distance is about 3,940 km (2,450 miles) β€” the Haversine baseline. Driving via the interstate network is roughly 4,500 km (2,800 miles), a circuity of about 1.14, taking around 40 hours of driving time. A typical flight covers about 4,130 km of air distance (great circle plus ~5%) in roughly 6 hours westbound. There is no direct passenger rail, but the historic rail routing exceeds 4,800 km. Same city pair, four defensible answers spanning 3,940 to 4,800+ km β€” which is exactly why 'distance' needs a qualifier.

Accuracy and limits

Every method has error bars. Great-circle math is exact for a sphere; the ~0.3% sphere-vs-ellipsoid difference is the only error, plus your coordinate precision. Driving distances depend on map freshness and route choice β€” expect Β±5% between engines and over time. Flight distances are estimates until the flight plan is filed. Rail distances are exact per the network but the network itself is what it is. And all methods inherit coordinate error: city 'locations' are usually city centres or airports, and measuring from the wrong point (downtown vs metro airport) can shift results by tens of kilometres. For contracts and quotes, state the method and the endpoints explicitly.

Which method should you use?

Match the method to the decision. Planning a road trip or estimating fuel: driving distance from a routing engine. Booking flights or estimating flight time: great-circle plus ~5%. Comparing how far apart places 'are' in the abstract, or doing geographic analysis: great-circle. Planning train travel: rail distance from the operator. Quick sanity check that two places are roughly X apart: great-circle in your head is fine. And when someone quotes you a distance without saying which kind, ask β€” the 20–50% spread between methods is where misunderstandings (and bad fuel budgets) live.

Measuring it yourself: tools and techniques

You rarely need to compute these by hand. For great-circle distance, any distance calculator that takes two lat/lng pairs and applies the Haversine formula gives the baseline in seconds β€” check that it reports kilometres, miles and ideally nautical miles. For driving distance, routing engines in map apps compute the road route; compare 'fastest' versus 'shortest' options, because they can differ noticeably. For flight distance, take the great-circle number and add about 5% for airways and climb/descent, then divide by cruise speed and add taxi time for a time estimate. For rail, the operator's timetable distance is authoritative. Whichever tool you use, sanity-check the result: if a driving distance comes out shorter than the great-circle distance, something is wrong β€” the road route can never beat the straight line.

Next steps

Run your own numbers in the distance calculator, compare with the driving distance calculator, or plan the whole journey with our road trip planning guide.

Frequently asked questions

Why is driving distance longer than the flight distance?

Roads cannot go in straight lines β€” they follow terrain, existing infrastructure, borders and cities. The ratio of road distance to straight-line distance (circuity) is typically 1.2–1.5. Aircraft fly the great-circle route directly, so flight distance stays close to the straight-line number.

Which distance should I use for fuel cost estimates?

Use driving distance from a routing engine for the route you will actually take, then apply your vehicle's real fuel efficiency. Great-circle distance will underestimate fuel needs by 20–50%.

What is circuity?

Circuity is the ratio of network distance (by road or rail) to straight-line great-circle distance. A circuity of 1.3 means the road route is 30% longer than the direct line. It measures how indirect the transport network is.

Can driving distance ever be shorter than the great-circle distance?

No β€” the great circle is the shortest possible surface path, so any road route is longer or (in theory) equal. If a tool reports a shorter driving distance, its endpoints or units are wrong.

Sources & data

Authoritative references used to research and verify this article:

🧭
Written by
JoseFitani
Editorial team

Articles and tools on JoseFitani are created and reviewed by the JoseFitani editorial team.

About JoseFitani β†’