This　paper　proposes　an　optimal　model　of　human　resource　allocation　at　the　planning　stage　of　software　projects.　Considering　the　different　human　resources　demands　required　by　different　tasks　and　the　characteristics　of　different　human　resources　roles,　this　model　realized　the　goal　of　assigning　ifferent　team　members　to　different　tasks　at　the　least　cost　under　the　conditions　of　satisfying　the　resource　demand　of　tasks　and　the　constraint　of　effort　of　different　stages,　which　is　based　on　accurate　estimation　by　the　information　available　at　the　planning　stage.　A　genetic　algorithm　is　then　designed　to　solve　this　complex　problem.　At　last,　an　example　is　given　to　verify　the　model.　The　results　show　the　feasibility　and　high　efficiency　of　the　model.　©　2010　IEEE.