In this paper, based on the full velocity difference model, and with the consideration of the difference between the driver’s psychological and actual headways, an extended heterogeneous car-following model is developed to investigate the effect of drivers’ different psychological headways. The analytical linear stable criterion for this extended heterogeneous traffic model is studied by using linear stability theory. The evolution of the heterogeneous traffic flow under a small perturbation is investigated by numerical simulations. The numerical and analytical results all show that the factor of psychological headway has important influence on the stability of the traffic flow. Different driver’s psychological headways have different effects on the traffic stability at different traffic densities. Compared with the full velocity difference model, the heterogeneous car-following model with smaller psychological headway can improve the stability at high traffic density, but the heterogeneous model with larger psychological headway lead to higher stability at the low traffic density.