Here, we report on the RAPD profiles of 61 individuals belonging to 10 macaque species. We used 23 different PCR primers on each sample and found an average of 17 bands per primer. The RAPD profiles appear to be highly reproducible because we found no differences in the amplification patterns produced by the DNA extracted from hair or blood. Strikingly, each species had a unique RAPD pattern homogeneously shared by all individuals. Comparisons between taxa showed that variability in the RAPD pattern was low, and the Sm index was below 0.601. Cluster analysis led to a division of the macaques into two main clusters: One withM. sylvanus and M. silenus and the other with M. arctoides, mulatta, fascicularis, nemestrina, tonkeana and fuscata. Macaca nigra and M. radiatawere positioned outside of these clusters. Gene flow may explain the zoogeographic pattern present in the RAPD profiles. The lack of within-species variability suggests the operation of founder effects and strong genetic drift, which may have been particularly strong in the case of peripherally placed species such asM. radiata, fuscata and nigra. The position of M. tonkeanais divergent from all the commonly accepted taxonomic and phylogenetic schemes. This result suggests that the RAPD technique is not always able to reveal the "true" phylogenetic relationships within the genusMacaca. The nature of genetic variation uncovered by the RAPD method is still unclear, and prudence should guide inferences about nucleotide divergence, population structure and phylogeny based solely on RAPD markers.