It　is　essential　to　know　drivers＇　learning　processes　and　decision　rules　in　actual　conditions　so　as　to　understand　their　choice　behavior　and　traffic　phenomenon.　This　paper　assumes　that　the　drivers　are　influenced　by　incomplete　traffic　information　in　addition　to　their　bounded　rationality.　Under　such　assumption,　this　paper　proceeds　to　construct　a　model　of　the　drivers＇　choice　behavior　on　the　basis　of　the　evolutionary　game　theory.　With　mathematical　induction　as　a　tool,　this　paper　proves　the　suitability　of　evolutionary　approach　for　the　analysis　of　drivers＇　route　choice　behavior.　It＇s　found　that　for　two　and　up　to　n　routes,　the　evolutionary　stable　state　derived　from　the　above-mentioned　approach　is　exactly　equivalent　to　equilibrium　derived　from　traffic　assignment　theory　and　stochastic　utility　theorem.