The possibility of identifying nonlinear time series using nonparametric estimates of the conditional mean and conditional variance is studied. Most nonlinear models satisfy the assumptions needed to apply nonparametric asymptotic theory. Sampling variations of the conditional quantities are studied by simulation and explained by asymptotic arguments for the first-order nonlinear autoregressive processes. The paper deals with the identification and prediction problems of the autoregressive models of nonlinear time series using nonparametric estimates of the conditional mean and conditional variance.