When allopatric species with incomplete prezygotic isolation come into secondary contact, the outcome of their interaction is not easily predicted. The parasitoid wasp Encarsia suzannae (iES), infected by Cardinium inducing cytoplasmic incompatibility (CI), and its sibling species E. gennaroi (EG), not infected by bacterial endosymbionts, may have diverged because of the complementary action of CI and asymmetric hybrid incompatibilities. Whereas postzygotic isolation is now complete because of sterility of F1 hybrid progeny, prezygotic isolation is still incipient. We set up laboratory population cage experiments to evaluate the outcome of the interaction between ES and EG in two pairwise combinations: iES vs EG and cured ES (cES, where Cardinium was removed with antibiotics) vs EG. We also built a theoretical model aimed at exploring the role of life-history differences and asymmetric mating on competitive outcomes. In three of four cages in each treatment, ES dominated the interaction. We found evidence for reproductive interference, driven by asymmetric mating preferences, that gave a competitive edge to ES, the species that better discriminated against heterospecifics. However, we did not find the fecundity cost previously shown to be associated with Cardinium infection in iES. The model largely supported the experimental results. The finding of only a slight competitive edge of ES over EG in population cages suggests that in a more heterogeneous environment the species could coexist. This is supported by evidence that the two species coexist in sympatry, where preliminary data suggest reproductive character displacement may have reinforced postzygotic isolation.