This paper develops a new method for the analysis of stochastic volatility (SV) models. Since volatility is a latent variable in SV models, it is dicult to evaluate the exact likelihood. In this paper, a non-linear ®lter which yields the exact likelihood of SV models is employed. Solving a series of integrals in this ®lter by piecewise linear approximations with randomly chosen nodes produces the likelihood, which is maximized to obtain estimates of the SV parameters. A smoothing algorithm for volatility estimation is also constructed. Monte Carlo experiments show that the method performs well with respect to both parameter estimates and volatility estimates. We illustrate the method by analysing daily stock returns on the Tokyo Stock Exchange. Since the method can be applied to more general models, the SV model is extended so that several characteristics of daily stock returns are allowed, and this more general model is also estimated. have investigated option pricing when volatility changes following a continuous stochastic process such as the Ornstein±Uhlenbeck process, and SV models are discrete approximations to such models.The maximum likelihood method is not easy to implement in SV models, but several alternative methods are now available (Ghysels, et al., 1996, andShephard, 1996 provide good reviews of the