This　paper　starts　with　the　geographical　lag　effect,　and　then　demonstrates　the　limitation　of　traditional　methods　about　the　forecasting　of　construction　land.　Through　adopting　the　related　software　to　verify　unary　regression　and　multiple　regressions　in　nonlag,　one　year　lag,　two　year　lag　and　three　year　lag　conditions,　and　see　the　other　social　and　economic　and　casual　factors　as　stochastic　disturbance　t.　So　we　can　exclude　the　worse　test　value　conditions,　which　reflect　the　degree　of　lag　and　its　state.　This　research　used　the　construction　land　and　social　and　economic　data　of　Huangpi　district,　Wuhan　City　from　1987　to　2004　to　analyze　the　lag　effect　in　land　forecast.　The　experiment　result　shows　us　the　construction　land　lags　the　social　and　economic　development　level,　which　actually　existing　objectively.　We　should　consider　the　influence　coming　from　time　lag　effect　when　we　forecast　the　amount　of　construction　land.　The　research　area＇　amount　of　construction　land　lag　social　and　economic　level　approximately　1-2　years.　The　forecast　result　which　is　under　lag　effect　eliminates　the　bad　effects　more.　©　2011　IEEE.