A mathematical model for the controlled degradation of polypropylene is presented in this article. A previous model of this process was extended to predict the whole molecular weight distribution of the modified resin. Probability generating functions were applied to transform the infinite set of mass balance equations of both polymer and radicals. The integration of the transformed set of equations yielded the probability generating function transforms. These transforms were then inverted with two different inversion algorithms, recovering the molecular weight distributions of the polymer. The model predictions were compared with our experimental data and other information taken from the literature. Good agreement was obtained. The approach presented here is also useful for other polymerization and postpolymerization processes. (C) 2003 Wiley Periodicals, Inc.