A　dynamic　Driving　simulation　experiment　of　thirty　subjects　is　carried　out　using　the　virtual　driving　simulator　to　analyze　the　variation　of　ECG　and　EEG　indicators　with　driving　time,　and　the　experiment　validates　the　effectiveness　of　the　ECG　and　EEG　indicators　as　the　evaluation　of　driving　fatigue.　There　is　a　significant　correlation　between　EEG　and　ECG　according　to　the　Pearson　Correlation.　Finally,　the　relationship　is　established　between　EEG　and　ECG,　and　the　overall　index　of　driving　fatigue　is　defined　by　Principal　component　analysis,　which　can　exclude　disturbed　factors　and　lower　data　volatility,　therefore,　to　improve　accuracy　of　driving　fatigue　evaluation.