In this paper various theories on traffic jams and stop-and-go wave modeling were reviewed and discussed. These include traditional microscopic car-following model, asymmetric traffic theory, first-order and second-order macroscopic models, traffic disturbance model, phase transition model as well as macroscopic simulation. The mechanism of stop-and-go waves, causes, generation, propagation, and absorption was discussed. A macroscopic simulation model was developed which suits for both free flow and stop-and-go traffic conditions. Ten hours of macroscopic simulation was performed on a section of freeway in California. The model predicted traffic speed and density were compared with field measured data. It was concluded that by applying proper boundary conditions for each roadway segment and incorporating capacity drop in the model, the macroscopic simulation model can reasonably predict the stop-and-go traffic states.