A set of 38 Leishmania stocks from the Andean valleys of Peru was characterized by both Multilocus Enzyme Electrophoresis (MLEE) and Random Amplified Polymorphic DNA (RAPD). Data were analyzed in terms of taxonomy and evolutionary genetics. Synapomorphic MLEE and RAPD characters, clear-cut clustering, and strong agreement between the phylogenies inferred from either MLEE or RAPD supported the view that Leishmania (Viannia) peruviana and Leishmania (Viannia) braziliensis correspond to two closely related, but distinct monophyletic lines (clades) and can therefore be considered as 'discrete typing units' (DTUs). The question whether the L. (V.) peruviana DTU deserves species status is dependent upon the desirability of it, in terms of epidemiological and medical relevance. A previous Orthogonal Field Alternating Gel Electrophoresis (OFAGE) analysis of the same L. (V.) peruviana isolates was published by Dujardin et al. (1995b). The data from the different markers (i.e. MLEE, RAPD and OFAGE) were compared by population genetics analysis. RAPD and OFAGE provided divergent results, since RAPD showed a strong linkage disequilibrium whereas OFAGE revealed no apparent departure from panmictic expectation. MLEE showed no linkage disequilibrium. Nevertheless, contrary to OFAGE, this is most probably explainable by the limited variability revealed by this marker in L. (V.) peruviana (statistical type II error). RAPD data were consistent with the hypothesis that the present L. (V,) peruviana sample displays a basically clonal population structure with limited or no genetic exchange. Disagreement between RAPD and OFAGE can be explained either by accumulation of chromosomal rearrangements due to amplification/deletion of repeated sequences, or by pseudo-recombinational events.