In a maple-beech forest in eastern New York, equal numbers of mounds, pits, and adjacent undisturbed soil microsites were censused for plant species density and cover over the growing seasons of 1978 and 1979. Several soil properties were also measured for each microsite. Detrended correspondence analysis (DCA) and chi-square goodness-of-fit tests were used to test whether the species were uniformly distributed over the microrelief positions. Comparisons of species composition, diversity, and vegetative phenologies were made between microsites and between hemlock-and hardwood-dominated forest areas. Differences in soil properties between microsites and between the two forest areas were assessed with one-tailed paired t tests and Wilcoxon rank-sum tests, respectively.Microrelief created a mosaic of soil properties and a vegetation pattern in the understory, both of which were affected by the presence of hemlock in the canopy. Ordinations (DCA) indicated two factors that affected species composition of the microsites: (I) presence of hemlock (Tsuga canadensis) as the nearest canopy tree, and (2) microtopography (whether the microsite was a mound, pit, or undisturbed-soil site). Species in non-hemlock areas had patchy distributions in the forest community: eight species grew mainly on mounds, six in pits, and five on undisturbed soil. The mounds and pits had characteristic assemblages of species. Species richness and total density and cover of plants in each microsite were constant over the growing season. In contrast, little spatial or temporal community pattern was apparent in forest areas influenced by hemlock.The effect of hemlock on species distributions was due in part to its effect on soil properties. For all microsites, forest areas with hemlock had greater soil organic matter content, available nitrogen, cation exchange capacity, and litter depth, lower soil calcium content, moisture content, temperature, pH, and A1 horizon depth; and no frost heaving.Microrelief also affected soil properties significantly. In comparison with pits, mounds were drier and poorer in nutrient content, and had a lower cation exchange capacity, less organic matter, less litter cover, a thinner A1 horizon, and less snow accumulation. In areas without hemlock, the mounds were also more acid, warmer in summer and colder in winter, and more subject to frost heaving than pits. Hemlock-influenced areas had no consistent microsite differences in pH or temperature.The different microenvironments, created by the interaction of microrelief and hemlock, resulted in patchy distributions of most understory species. This pattern was likely a result of species requirements for and tolerances of environmental conditions, tempered by competitive interactions.