The performance of different molecular markers in the assessment of population structure was tested using samples of Solea vulgaris collected in the Mediterranean within and outside the hypothetical dispersal ability of the species. A total of 172 individuals belonging to four population samples were analysed using 15 microsatellites [simple sequence repeats (SSRs)] and 153 amplified fragment length polymorphisms (AFLPs). Considering the global qualitative patterns, we found a correlation between SSRs and AFLPs in detecting genetic differentiation among samples. However, on a small geographical scale, AFLPs were able to discriminate individuals from neighbouring populations whereas SSRs were not, and the percentage of individuals correctly assigned to their population of origin was higher with AFLPs than with SSRs. The high number of loci analysed with the AFLP technique could increase the probability to include outlier loci in the analysis; however, the neutrality test performed on our data set did not show evidence of selection acting on the S. vulgaris samples. Even if the choice of the molecular marker depends mainly on the biological question to be addressed, the higher power of discrimination and the comparative technical ease of obtaining data from AFLPs with respect to SSRs suggest the use of AFLPs for many population genetics studies.