The aim of this study was to evaluate two methods of nucleic acid extraction (spin-column-based method - commercial kit and direct boil - DB) from swab sampling compared to biopsy sampling for the diagnosis of tegumentary leishmaniasis (TL), (cutaneous - CL and mucocutaneous - MCL forms). The impact of these nucleic acid extraction protocols on different types of PCR and LAMP techniques were compared regarding nucleic acid quality, molecular assays accuracy, indirect quantitation, and costs. The evaluated patients were 57 TL cases (36 CL and 21 MCL) and 34 non-cases. Swab samples extracted by the DB method showed a higher DNA degradation rate and worse DNA quality in comparison to the commercial kit. Molecular tests performed on biopsy samples showed identical or higher performance in all analysis, as compared to their own performance on swab samples for TL (CL and MCL). However, only the SSU rRNA TaqMan™ RT-PCR test showed a significant difference between the performance of biopsy and swab samples extracted by commercial kit. The kDNA-cPCR coupled with swab extracted by commercial kit showed the highest accuracy (95.6%) for TL diagnosis. The sensitivity of the LAMP-RT 18S method in swab samples extracted with a commercial kit (82.5%) was close to that found in biopsy samples (86%) for TL diagnosis. The DB extraction method presented the lowest cost. The use of swab as a minimally-invasive sampling method, associated with an efficient nucleic acid extraction protocol, may represent a low-cost alternative for the diagnosis of CL and MCL.