Various RR techniques for estimating the proportion of individuals belonging to a sensitive category were presented in the previous chapter. In this chapter, such procedures are extended to the situation where the population is classified into several disjoint categories, at least one of which is sensitive and at most one is non-sensitive. Most of the polychotomous techniques are based on the applications of Warner, Unrelated Question, Forced RR, and their combinations. Details of a few important randomization techniques and methods of estimating population proportions from such techniques are discussed here.