Decision-making　model　of　queuing　and　queue-jumping　of　drivers　at　intersections　is　proposed　concerning　the　behavior　of　queue-jumping.　The　evolutionary　process　that　drivers　choose　strategies　of　queuing　and　queue-jumping　is　analyzed　using　evolutionary　game　theory.　A　duplicative　dynamic　equation　is　established　and　evolutionary　stable　state　is　analyzed.　The　results　show　that　the　payoff　of　queuing　drivers　is　larger　than　that　of　queue-jumping　drivers　when　the　queue-jumping　drivers　of　violating　regulations　can　be　timely　found　and　punished　by　the　traffic　management　department.　The　strategy　of　queuing　is　the　evolutionary　stable　state.