Based on the theory of random vibration, a quarter car model is used in this research for the prediction of ride comfort for passenger cars. Authenticity of analytical results is then examined by correlation analyses with direct measurements and human evaluation on a test track, respectively. The approach indicates that good correlation exists between simulation results and RMS values from the measurements at high speeds. However, analytical results are not correlated well with power spectral density functions, which can be improved by using nonlinear suspension damping model in the approach. On the other hand, the original acceleration and ISO weighted acceleration indices developed in this project are found both capable of predicting the proper trend of human subjective evaluation.