BACKGROUND: Sand fly species able to support the survival of the protozoan parasite Leishmania have been classified as permissive or specific, based upon their ability to support a wide or limited range of strains and/or species. Studies of a limited number of fly/parasite species combinations have implicated parasite surface molecules in this process and here we provide further evidence in support of this proposal. We investigated the role of lipophosphoglycan (LPG) and other phosphoglycans (PGs) in sand fly survival, using Leishmania major mutants deficient in LPG (lpg1(-)), and the phosphoglycan (PG)-deficient mutant lpg2(-). The sand fly species used were the permissive species Phlebotomus perniciosus and P. argentipes, and the specific vector P. duboscqi, a species resistant to L. infantum development. PRINCIPAL FINDINGS: The lpg2(-) mutants did not survive well in any of the three sand fly species, suggesting that phosphoglycans and/or other LPG2-dependent molecules are required for parasite development. In vitro, all three L. major lines were equally resistant to proteolytic activity of bovine trypsin, suggesting that sand fly-specific hydrolytic proteases or other factors are the reason for the early lpg2(-) parasite killing. The lpg1(-) mutants developed late-stage infections in two permissive species, P. perniciosus and P. argentipes, where their infection rates and intensities of infections were comparable to the wild type (WT) parasites. In contrast, in P. duboscqi the lpg1(-) mutants developed significantly worse than the WT parasites. CONCLUSIONS: In combination with previous studies, the data establish clearly that LPG is not required for Leishmania survival in permissive species P. perniciosus and P. argentipes but plays an important role in the specific vector P. duboscqi. With regard to PGs other than LPG, the data prove the importance of LPG2-related molecules for survival of L. major in the three sand fly species tested.