Population genetic analyses of the mycorrhizal fungus Cenococcum geophilum were conducted to test for a clonal or recombining population structure. Multilocus genotypes based on polymorphisms in 9 loci, identified in this study by PCR-SSCP techniques, were obtained for two populations. Genotypic variation occurred on a fine scale because unique genotypes were identified at most every transect point, and in some cases occurred even within one soil sample (equivalent to about a 500 mL volume). The largest genet observed occurred over a 30 meter transect space. The two population genetic methods employed to distinguish between clonality and recombination, (1) Index of Association; and (2) "Parsimony Tree Length Permutation Test" (PTLPT), could not reject the null hypothesis of recombination in either population. Wright's Fst, as estimated by theta, was used to examine gene flow between the two populations based on allele frequencies. Two of the nine loci had theta values that were not significantly different from what one would expect for the null hypothesis of panmixia. However, the other seven loci were consistent with reduced gene flow. The theta value for the Fisher combined probability (combining all 9 loci) was significant and indicated that there was genetic differentiation between these two populations.