Does plant diversity drive soil microbial diversity in temperate, upland grasslands? Plants influence microbial activity around their roots by release of carbon and pot studies have shown an impact of different grass species on soil microbial community structure. Therefore it is tempting to answer yes. However, evidence from field studies is more complex. This evidence is reviewed at three different scales. First, studies from the plant community scale are considered that have compared soil microbial community structure in pastures of different vegetation composition, as a consequence of pasture improvement. These show fungi dominating the biomass in unimproved pastures and bacteria when lime and fertilizers have been applied. Secondly, evidence for interactions between individual grass species and soil microbes is discussed at the level of the rhizosphere, by considering both pot experiments and field studies. These have produced contrasting and inconclusive results, often due to spatial heterogeneity of soil properties and microbial communities. In particular, increased soil pH and fertility in urine patches and other nutrient cycling processes interact to increase the spatially complexity of soil microbial communities. Finally three studies which have measured microbial community structure in the rhizoplane are considered. These show that bacterial diversity is not directly related to plant diversity, although fungal diversity is. In addition, the soil fungal community has been demonstrated to have an effect upon the composition of the bacterial community. We suggest that while current vegetation influences fungal communities (particularly mycorrhizae) and litter inputs fungal saprotrophs, bacterial community structure is influenced more by the quality or composition of soil organic matter, thereby reflecting carbon inputs to the soil over decades.