Abstract This project is part of work underway in our laboratories to test the hypothesis that the induction of soil suppressiveness to plant parasitic nematodes that occurs following planting of velvetbean (Mucuna deeringiana (Bort) Merr.) is associated with the development of an antagonistic microflora in soils and rhizospheres. The specific objective of this investigation was to examine long-term microbial shifts associated with the use of velvetbean in rotations to control nematodes. A crop rotation study was conducted in microplots, consisting of three crop cycles. Cycle 1 involved planting of either velvetbean or cowpea (Vigna unguiculata L.) in the first spring. Cycle 2 during the next fall and winter was fallow or cover-cropped with wheat (Triticum aestivum L.) or crimson clover (Trifolium incarnatum L.). Cycle 3 the next spring was soybean (Glycine max (L.) Merr.). Populations and species diversity of bacteria and fungi in soils or rhizospheres were investigated at the end of each cropping cycle. Rhizosphere fungal populations were significantly smaller on velvetbean than on cowpea at the end of cycle 1. The use of velvetbean in cycle 1 significantly decreased rhizosphere bacterial populations on crops in cycle 2, compared to treatments which had cowpea in cycle 1. Velvetbean also influenced bacterial diversity, generally increasing frequency of bacilli, Arthrobacter spp. and Burkholderia cepacia, while reducing fluorescent pseudomonads. Some of these effects persisted through cycle 3. Fungal diversity was influenced in cycle 1 by velvetbean; however, effects generally did not persist through cycles 2 and 3. The results indicate that the use of velvetbean in a cropping system alters the microbial communities of the rhizosphere and soil, and they are consistent with the hypothesis that the resulting control of nematodes results from induction of soil suppressiveness.