A computational method for simulation of stochastic processes of an electron multiplication in microchannel electron amplifiers is developed. The method is based on 3D Monte Carlo (MC) simulations and theorems about serial and parallel amplifcation stages proposed here. Splitting a stochastic process into a number of di®erent stages, enables a contribution of each stage to the entire process to be easily investigated. The method preserves all advantages of the MC simulations which are used only once for one simple stage. The use of the theorems allows to conduct any further investigations and optimizations without additional MC simulations. The method provides a high calculation accuracy with minimal cost of computations. The model is relevant to unsaturated operation of the amplifier with ex- ponential output distribution. The mean gain and the variance of the amplitude distribution at the output of the amplifier are predicted. In this paper the method is used to show how the input ratio of the signal to the noise is transforming to the output one, and how differ- ent multiplication stages contribute to the noise factor of the system. Finally, the effect of variations in channel diameters on noise characteristics of microchannel electron amplifiers is investigated.