On strongly acidic sandy clay loams in southeastern Australia there can be problems with the growth and persistence of subterranean clover in clover-ley rotations. During a 4 year rotation, which involved wheat crops and then the re-establishment of clover, the effects of direct drilling and cultivation, and agronomic methods on the nodulation of subterranean clover and the survival of Rhizobium leguminosarum biovar trifolii were compared. In the season in which the clover was sown, the plants on the unlimed soil were poorly nodulated, with no nodules at all on the clover that was sown without inoculation and where the soil had been cultivated. Inoculation of the clover seed increased the nodulation of the clover in the absence of lime. The populations of R. trifolii were estimated throughout the wheat-clover sequence in the rotation. Prior to establishing the experiment the site had been in a subterranean clover-based pasture for 12 years, and the populations of rhizobia were about 200–300 g −1 soil. Without lime the rhizobia populations declined to about 10 or less rhizobia g −1 soil over the period the site was cropped, but where the soil was limed there was no change in the rhizobia populations with cropping. When the subterranean clover was re-established the populations of R. trifolii were increased with all soil treatments, with the unlimed soil obtaining populations similar to those present at the start of the experiment and the rhizobia number on the limed soil increasing to about 10 5 g −1 soil. The most favoured soil environment for R. trifolii was where the pH was raised mostly in the soil surface, and the soil was only minimally disturbed with the direct-drill treatment.