BackgroundThe Sordida subcomplex (Triatominae) comprises four species, Triatoma garciabesi, T. guasayana, T. patagonica and T. sordida, which differ in epidemiological importance and adaptations to human environments. Some morphological similarities among species make taxonomic identification, population differentiation and species delimitation controversial. Triatoma garciabesi and T. sordida are the most similar species, having been considered alternatively two and a single species until T. garciabesi was re-validated, mostly based on the morphology of male genitalia. More recently, T. sordida from Argentina has been proposed as a new cryptic species distinguishable from T. sordida from Brazil, Bolivia and Paraguay by cytogenetics. We studied linear and geometric morphometry of the head, wings and pronotum in populations of these species aiming to find phenotypic markers for their discrimination, especially between T. sordida and T. garciabesi, and if any set of variables that validates T. sordida from Argentina as a new species.ResultsHead width and pronotum length were the linear variables that best differentiated species. Geometric morphometry revealed significant Mahalanobis distances in wing shape between all pairwise comparisons. Triatoma patagonica exhibited the best discrimination and T. garciabesi overlapped the distribution of the other species in the morphometric space of the first two DFA axes. Head shape showed differentiation between all pairs of species except for T. garciabesi and T. sordida. Pronotum shape did not differentiate T. garciabesi from T. guasayana. The comparison between T. garciabesi and T. sordida from Argentina and T. sordida from Brazil and Bolivia revealed low differentiation based on head and pronotum linear measurements. Pronotum and wing shape were different between T. garciabesi and T. sordida from Brazil and Bolivia and T. sordida from Argentina. Head shape did not differentiate T. garciabesi from T. sordida from Argentina.ConclusionsWing shape best delimited the four species phenotypically. The proposed cryptic species, T. sordida from Argentina, differed from T. sordida from Brazil and Bolivia in all measured shape traits, suggesting that the putative new species may not be cryptic. Additional studies integrating cytogenetic, phenotypic and molecular markers, as well as cross-breeding experiments are needed to confirm if these three entities represent true biological species.Electronic supplementary materialThe online version of this article (10.1186/s13071-017-2350-y) contains supplementary material, which is available to authorized users.