The basic assumption of the Serber model in the description of high- energy nuclear reactions is that the interactions of incident particles with complex nuclei can be described in terms of individual particle-particle collisions within the nucleus. Calculations were performed making use of the basic assumption of the Serber model, a more realistic nuclear model, recent cross-section data, and an exact statistical sampling technique. The sampling technique has not been used previously in calculations of this type. Calculations were performed for incident e (as Fe/sup 59 /sup +/, e (as Fe/sup 59 /sup -/, neutrons, and protons on nuclei from lithium to uranium. The energy range of the incident particles varied from about 50 to 350 Mev, i.e., the energy region in which pion production is not likely. Free-particle cross sections were used in determining the collisions within the nucleus, and statistical sampling techniques were used throughout, The problem was coded for the IBM-7090. Extensive comparisons with experiment were made and the results indicate that the calculation can be used to predict most of the cascade data for incident nucleons on complex nuclei, but only the gross features of the data are predictable for incident pions on nuclei. The effects …