The Index of Association (IA) has been proposed by Maynard Smith et al. (1993) as a general method for characterizing the population structures of microorganisms as either: clonal, epidemic, cryptic species or panmictic. With reference to the current debate surrounding the mode of reproduction in parasitic protozoa, this study explores (i) the suitability and limitations of the IA for characterizing populations of Trypanosoma brucei s.l., and (ii) the idea that the significance of genetic differences between populations may be better understood if the evolution, spread and temporal stability of certain parasite genotypes are also considered. Four populations of T. brucei from Côte d'Ivoire, Uganda and Zambia are analysed using the IA and a complementary test for linkage disequilibrium, test f of Tibayrenc, Kjellberg & Ayala (1990). The two populations from Uganda are characterized as epidemic, while the others appear more or less clonal; the merits of the two methods are compared. The implications of the various population classifications are discussed with reference to genotype longevity in each region; the evolutionary and biomedical consequences of the genetic non-homogeneity of T. brucei are reviewed.