If the H-2 congenic mouse strains A.SW (H-2n) and A.CA (H-2f), are infected with Trypanosoma cruzi, a 45 kDa protein (Tc45), present in cultured epimastigotes and blood trypomastigotes, is recognized only by the A.SW strain sera. In order to explore the possibility that among seropositive humans the response to Tc45 is also highly variable, 81 chagasic human sera (as defined by the HemAve agglutination test, Polychaco S.A.I.C., Buenos Aires, Argentina) were tested in a direct (epimastigote antigenic complex directly bound to the solid phase) and indirect immunoradiometric assay (IRMA) (Tc45, from a partially purified preparation, bound to the solid phase, by means of a monoclonal antibody). Sixty nine of these sera reacted in both the direct and indirect assays, 11 were negative in both assays (these samples may correspond to false positives detected by the commercial agglutination test) and only one reacted with the antigenic complex but not with Tc45. Reactivity of the human sera with the epimastigote antigenic extract was relatively homogenous, while reactivity with Tc45 was extremely variable. No statistical correlation was determined between the two variables. Given the high variability of the human response to Tc45, ranging from negative to highly positive, together with the immunogenetic restriction previously described in the murine model, we speculate that human MHC may also modulate the response to this molecule.