Maximum likelihood technique 1s used to estimate the parameters of single input single output nonlinear systems. Two algorithms are described and applied . One is for the identification of Hammerstein nonlinear models, which is useful if no priori knowledge about the mathematical form of the nonlinearity is available. The other algorithm is for the identification of systems which have known forms for the nonlinearities. It is derived for continuous nonlinear systems, and applied for simulated data generated from linear and nonlinear second order continuous models. It is also used to fit linear, and nonlinear second order continuous models to practical data taken from a test on the glucose homeostatic control system of dogs. The emphasis is on obtaining simplified algorithm for continuous nonlinear systems in order to save computing time, and get satisfactory results.