Abstract: In this paper, we apply the theory of the modern martingales and stochastio integralsto study the strong consistency of maximum likelihood estimate for stochastic processes, and get some sufficient conditions. Also, we generalize the work of P. E. Caines （1975） on the ease of a finite paramour set.