The sensitivity and reproducibility of a PCR targeted to amplify the conserved 120 base-pair region of minicircles from Leishmania kDNA was defined using DNA extracted from skin biopsy imprints on filter paper. Seventy-seven patients with cutaneous leishmaniasis from an endemic region of Leishmania (Viannia) braziliensis in Brazil underwent skin biopsy of the ulcer border. Tissue samples were imprinted on filter paper and then, they were stored at -20 degrees C. Imprints on filter paper were stored at 4 degrees C. Samples were processed at three laboratories; Lab1 and Lab2 performed the PCR-kDNA assay using DNA extracted from the filter paper, and Lab3 processed PCR-kDNA using DNA from fresh-frozen tissue used as a gold standard. All samples were codified to maintain blinding during lab processing. Fifty-three (68.8%) patients had parasites isolated and identified by isoenzymes as L. (V.) braziliensis. The positivity of PCR-kDNA was similar between the three laboratories: 87.0, 85.7 and 88.3% (Lab1, Lab2 and Lab3, respectively). The sensitivity of PCR-kDNA in culture-proven cases was better, and showed similar results in all laboratories: 95.8, 95.8 and 97.9% (Lab1, Lab2 and Lab3, respectively). Data from the 77 enrolled patients showed an overall percent agreement of 80.5% (Kappa=0.173) for the filter-paper approach between Lab1 and Lab2. Percent agreement between Lab1 and Lab3 was 83.1% (Kappa=0.22), and it was 94.8% between Lab2 and Lab3 (Kappa=0.77). Fifteen patients were diagnosed in just one of the two laboratories that used DNA extracted from filter paper. We conclude that the sensitivity of the filter paper approach is satisfactory and could be used in clinical trials and field work. Reproducibility could be improved using two separate imprints from the same biopsy sample.