Abstract Studies on experimental sand fly infection require the availability of colonies and laboratory conditions. In Brazil, Lutzomyia longipalpis (Diptera: Psychodidae) (Lutz and Neiva 1912) is responsible for the highest infection rates by Leishmania spp. and this species is one of the most suitable species for laboratory colonization. In this study, we describe a method for growing Lu. longipalpis in laboratory conditions (10 generations) from natural populations sampled from a region of high endemicity for visceral leishmaniasis in the state of Maranhão, Northeastern Brazil. Using two methods (individualized or grouped females), the colony’s highest productivity occurred in the first four generations, where all stages presented with high frequency. Nonviable eggs represented more than 50% of the total eggs produced by engorged females, while pupae were more resistant to fungal contamination, with a mortality rate of only 2%. In both methods, there was a predominance of female emergence; however, the ratio between males and females did not show significant differences, IF (P = 0.8023) and GF (P = 0.1831). Using the method of individualized females, the F4 generation took the longest to appear (234 d; 64 ± 57 d); by grouped females, F3 took the longest to appear (102 d; 47 ± 20 d). This method provides sufficient numbers of insects to perform vector competence tests for Leishmania spp. that cause the cutaneous form of leishmaniasis, usually found in Lu. longipalpis sampled from the study location.