Evolution of microsatellites (or simple sequence repeats, SSRs) is a complex process that converts perfect repeats to novel structural elements with functions poorly understood, such as imperfect and compound microsatellites. An in silico analysis often Burkholderia pseudomallei genomes revealed 215683 micro-satellites, and more than 98% of them proved imperfect. The density of microsatellites in the genome ranged from 2922.7 to 3022.6 per Mbp. Approximately 10.20-10.67% of the repeats were parts of compound micro-satellites. The of compound microsatellite density varied from 144.7 to 150.6 per Mbp. Between-strain differences in microsatellite distribution were explained by a direct correlation of the SSR density with the GC content and an inverse relationship between the SSR density and the genome size. For each B. pseudomallei chromosome, the SSR density similarly correlated with its size and GC content. Chromosome 2 showed a significant correlation between the SSR and compound microsatellite densities (r = 0.93, p < 10^-3). The association of imperfect and compound microsatellite densities with the structural features of each chromosome and the fact that motifs are degenerate and occur in few copies in the majority of B. pseudomallei microsatellites agree with the previous hypothesis of negative selection affecting extended SSRs. The mechanism of selection possibly involves an accumulation of point mutations, which lead to an interruption of the repeat during replication because easily passable secondary structures may form to stabilize the microsatellite length.