During the 1980s, a large number of typing methods for the strain differentiation of Candida albicans were described in the literature. Although these methods have been based on a variety of physiological and genetic markers, none is ideal. This review discusses the characteristics of an ideal typing method in terms of its typability, reproducibility, and discriminatory power. Ways of determining these characteristics are presented so that the available typing methods for Candida albicans can be objectively compared. Available typing methods for C. albicans include serotyping, morphotyping, resis-totyping, biotyping, and killer yeast typing. Electrophoretic methods include immunoblotting, isoenzyme analysis, analysis of DNA restriction fragment length polymorphism, karyotyping, and the use of DNA probes. The application of these methods to epidemiological research, the investigation of outbreaks of disease, and die study of virulence is described. The potential impact of the phenomenon of phenotypic switching on the reproducibility of these typing methods is discussed. It is concluded that many of the available typing methods have not been adequately assessed by their developers and that several have only poor discriminatory power or reproducibility.