Genotypic variation among 62 strains of Aphanomyces euteiches, four of A. cochlioides, and a Saprolegnia sp. was investigated using RAPD analysis. Pathogenicity assays on pea, bean, alfalfa, red clover, and sugarbeet were used to determine host preference among the strains of A. euteiches and A. cochlioides. Pathogenicity tests revealed six pathotypes of A. euteiches with host preferences for bean, alfalfa, pea, pea and alfalfa, red clover and alfalfa, and bean and alfalfa. Another group of strains was non-pathogenic on the five plant species. The host of origin tended to be the host on which each strain incited the highest disease severity. A. euteiches did not incite root rot symptoms on sugarbeet, and A. cochlioides was pathogenic only to sugarbeet. RAPD analyses provided a measure of genetic diversity in Aphanomyces. Fifty random decanucleotide primers were screened with five test strains from four hosts representing different pathotypes, and 32 primers amplified DNA fragments from all five strains. Eight primers were chosen for most of this study based on number and polymorphic nature of the bands generated. RAPD assays of 62 strains of A. euteiches with the eight primers yielded 159 polymorphic and no monomorphic, strongly amplified bands. Cluster analyses of RAPD data revealed genotypic differences among three groups of A. euteiches which corresponded to their host of origin and host preference for bean, alfalfa, and red clover/alfalfa. Strains nonpathogenic on all plants tested formed another genotypic group, corroborating results from the pathogenicity assays that indicated this is a discrete group. The bean and non-pathogenic groups were the most distinct. The A. cochlioides and Saprolegnia strains were genotypically distinct from the pathogenic, but not the non-pathogenic strains of A. euteiches. Reproducibility of RAPD assays was confirmed by replicated amplifications and DNA hybridization analysis. Results indicated that A. euteiches is composed of distinct subspecific groups based on genotype and host preference.