The causative agent of potato brown rot and bacterial wilt, Ralstonia solanacearum, results in serious world-wide economic losses, particularly in the tropics. In the last decade, however, the incidence of bacterial wilt in potatoes grown in Northern Europe has increased, presenting an interesting epidemiological puzzle. Its occurrence may be as a result of changes in agricultural practice or the emergence of a novel bacterial variety, better adapted to cooler conditions. To understand the distribution and genetic diversity of this phytopathogen, we have analysed a collection of 82 isolates from Europe and tropical regions. Both phenotypic [SDS-PAGE (sodium dodecyl sulphate polyacrylamide gel electrophoresis) profiling, FAME (fatty acid methyl esters) analysis, growth profiles and EPS (exopolysaccharide) production] and genotypic [16S rRNA RFLP (restriction fragment length polymorphism), ARDRA (amplified ribosomal DNA restriction analysis) and sequence analysis of 16S-23S rRNA ITS and flanking regions] methods were compared. Principal component analysis of FAME profiles clustered isolates into three groups and ARDRA of a 0.85 kb amplified fragment from the 16S-23S ITS region differentiated isolates into four groups. Using sequence analysis, specific primers were designed within the variable region 147-170 of the 23S rRNA. These primers, RsolT2 and RsolT3, respectively, differentiated isolates into two distinct clusters as described previously by Wullings and colleagues (Wullings et al., 1998). The European strains (Biovar 2, race 3) analysed in this study specifically hybridized with RsolT3, and showed considerable genetic homogeneity when compared with strains of other races from 'the rest of the world'. These data indicate the possible selection and proliferation of a 'European'-adapted variant.