A random sampling technique in which polymer molecules are sampled from an infinite number of polymer molecules in the reaction mixture is used to predict the molecular weight distribution (MWD) change during random scission reactions of star-shaped polymers. For simpler initial MWDs, the analytical solutions for the full MWD change, as well as the average molecular weights, can be derived in a straightforward manner. For any initial MWD, the concept of the random sampling technique provides a unique Monte Carlo simulation algorithm that enables one to estimate the change in the statistical properties during random degradation quite effectively. Compared with the random degradation of linear polymer molecules, the decrease in the weight-average molecular weight is slower; however, such differences are made smaller as the initial MWD becomes broader. In the course of random degradation, linear and branched polymer fractions are formed. As the degradation reaction proceeds, the polydispersity index of the linear and branched fractions approaches 2 and (f + 1)/f, respectively, irrespective of the initial polymer distribution, where fis the number of arms in a star polymer.