In developing countries, various plans are used to contain urban sprawl of large cities, though they are rarely successful. Previous studies mainly attribute the failure to unsystematic urban planning, problems of governance or implementation, and rapid economic development. We suggest that an underestimation of urban population numbers is also a key factor, and that population forecasting methods currently in use need to be revised. Using panel data covering 110 major countries across 60 years (1950–2010), this study examined the relationship between large cities, and countries' population numbers, urbanization rate, land area, and four dummy variables. Based on this, a simple and generalizable model is developed to project the population of large cities.