Measuring Urban Area is messier than you think
The biggest problem people run into when ranking cities by area is that they never specify which measurement they're using. Administrative boundaries, urban footprint, and metropolitan area are three entirely different numbers, and they produce wildly different results. I spent months reconciling these datasets for a comparative geography project, and the friction is real. Different sources define their boundaries differently, population counts get revised on different schedules, and some municipalities aggressively annex surrounding rural land to inflate their stats. If we're talking strictly about municipal administrative boundaries, the ranking looks very different than if we measure actual continuous urbanized land. Chongqing, China tops the list at roughly 82,400 square kilometers, which sounds enormous but includes vast rural mountains and farmland within its jurisdiction. By most standard rankings that use built-up urban area, the second-largest city tends to be either Shanghai or sometimes Guangzhou depending on the dataset, with Shanghai's continuous urban footprint sitting around 6,400 to 7,000 square kilometers in recent measurements. Here's where it gets complicated and most people get it wrong. The Greater Bay Area metro cluster and the Beijing-Tianjin-Hebei megalopolis blur the lines significantly. Tianjin alone covers about 11,900 square kilometers administratively, which would put it ahead of Shanghai in raw jurisdictional area, but its actual dense urban core is smaller. I hit this exact wall when cross-referencing multiple sources and found that some datasets count satellite towns within commuting distance as part of the urban area while others draw a hard line at the last continuous built-up zone. The workaround I used was pulling from the Global Human Settlement Layer's built-up area raster data, which gives a more consistent picture of actual constructed land regardless of political boundaries, though it requires downloading and processing geospatial files rather than just looking up a number.
A counter-intuitive point that catches people off guard: larger administrative area often correlates with lower urban quality of life metrics. Cities with massive jurisdictions tend to have enormous rural populations within their borders that drag down per-capita infrastructure investment. Chongqing's administrative area is comparable to some small countries, yet its urbanization rate within those boundaries is closer to 70 percent than 95 percent, meaning nearly a third of its official area is just countryside and mountains not contributing to the urban character anyone would recognize. Another thing beginners consistently overlook is how frequently city boundaries expand. Chinese municipalities have been annexing surrounding counties for decades, so a ranking from 2015 will differ meaningfully from one based on 2024 data. Indian cities follow a similar pattern with municipal corporation expansions. This means any ranked list has a built-in expiration date unless you're tracking boundary changes alongside area figures. Brazilian cities like Manaus and Belém also appear high on some lists due to enormous forested administrative zones, but their actual urban footprints are modest by comparison. The practical limitation everyone needs to accept is that there is no single authoritative global database for this. The Demographia World Urban Areas report uses continuous built-up land and is probably the most respected methodology, but it doesn't cover every city and sometimes lags behind current boundary changes. The World Population Review and similar aggregators pull from various national census sources with inconsistent definitions, so their rankings are unreliable for serious work. If you need accuracy for a research project or professional use, plan to pull raw data from each country's census bureau and process it yourself rather than trusting a compiled list. That process typically takes a weekend for five to ten cities and eliminates the kind of discrepancies that come from layering one aggregator's errors on top of another's.
For casual reference purposes, the second-largest city by urban built-up area in most modern datasets is Shanghai, followed closely by Guangzhou-Foshan if you count the contiguous metropolitan zone. If you're using administrative boundaries instead, then cities like Harbin at roughly 53,000 square kilometers or Istanbul at around 5,400 square kilometers appear differently depending on which countries you include and how you handle transcontinental definitions.
Get the Full Details
