Microsatellite markers were used to investigate the genetic structure among invasive L. decolor populations from Australia and a single international population from Kansas, USA to determine patterns of dispersal. Six variable microsatellites displayed an average of 2.5-4.2 alleles per locus per population. Observed (H(O)) heterozygosity ranged from 0.12-0.65 per locus within populations; but, in 13 of 36 tests, H(O) was less than expected. Despite low levels of allelic diversity, genetic structure estimated as theta was significant for all pairwise comparisons between populations (theta = 0.05-0.23). Due to suspected null alleles at four loci, ENA (excluding null alleles) corrected F(ST) estimates were calculated overall and for pairwise population comparisons. The ENA-corrected F(ST) values (0.02-0.10) revealed significant overall genetic structure, but none of the pairwise values were significantly different from zero. A Mantel test of isolation by distance indicated no relationship between genetic structure and geographic distance among all populations (r(2) = 0.12, P = 0.18) and for Australian populations only (r(2) = 0.19, P = 0.44), suggesting that IBD does not describe the pattern of gene flow among populations. This study supports a hypothesis of long distance dispersal by L. decolor at moderate to potentially high levels.