This　research　examines　the　spatio-temporal　distribution　of　population　in　an　urban　environment　and　its　relationship　with　urban　functions　using　an　unprecedented　high-resolution　and　broad-coverage-crowd　LBS　dataset　from　Tencent,　one　of　the　biggest　Internet　companies　in　China.　By　examining　the　distribution　of　the　population　during　different　time　periods,　different　urban　morphologies　are　observed.　The　analysis　of　the　spatio-temporal　population　distribution　based　on　temporal　entropy　indicates　that　population　distributions　of　employment,　commercial,　and　scenic　areas　have　larger　temporal　fluctuations　than　those　in　residential　and　mixed-use　areas.　As　for　the　Spearman　correlation　coefficient　between　urban　functions　and　temporal　population　distribution　based　on　300 × 300-m　grids,　it　is　then　measured　to　uncover　the　underlying　reason　for　this　spatio-temporal　distribution　of　population.　The　result　demonstrates　that　as　urban　functions　become　more　mixed,　the　temporal　distribution　of　the　population　becomes　more　even.　At　a　local　scale,　temporal　population　distribution　in　key　areas　shows　that　the　location　of　people　in　a　certain　place　is　in　accordance　with　human　behavior.　The　population　in　employment-dominated　areas　shows　large　fluctuations　on　weekdays　but　is　relatively　evenly　distributed　on　weekends.　The　population　in　commercial　areas　only　peaks　for　several　hours　on　both　weekdays　and　weekends.　Comparatively,　mixed　areas　and　large-scale　residential　communities　accommodate　a　stable　number　of　people　at　all　times.　©　2016,　Springer　Science+Business　Media　Dordrecht.