A glasshouse experiment was undertaken to provide baseline data on the variation between conventional maize (Zea mays L.) varieties and genetically modified maize plants expressing the insecticidal Bacillus thuringiensis protein (Bt, Cry1Ab). The objective was to determine whether the variation in soil parameters under a range of conventional maize cultivars exceeded the differences between Bt and non-Bt maize cultivars. Variations in plant growth parameters (shoot and root biomass, percentage carbon, percentage nitrogen), Bt protein concentration in shoots, roots and soil, soil nematode abundance and soil microbial community structure were determined. Eight paired varieties (i.e. varieties genetically modified to express Bt protein and their near-isogenic control varieties) were investigated, together with a Bt variety for which no near-isogenic control was available (NX3622, a combined transformant expressing both Bt and herbicide tolerance) and a conventional barley (Hordeum vulgare L.) variety which was included as a positive control. The only plant parameter which showed a difference between Bt varieties and near-isogenic counterparts was the shoot carbon to nitrogen ratio; this was observed for only two of the eight varieties, and so was not attributable to the Bt trait. There were no detectable differences in the concentration of Bt protein in plant or soil with any of the Bt-expressing varieties. There were significant differences in the abundance of soil nematodes, but this was not related to the Bt trait. Differences in previously published soil nematode studies under Bt maize were smaller than these varietal effects. Soil microbial community structure, as determined by phospholipid fatty acid (PLFA) analysis, was strongly affected by plant growth stage but not by the Bt trait. The experimental addition of purified Cry1Ab protein to soil confirmed that, at ecologically relevant concentrations, there were no measurable effects on microbial community structure.