The present study compares the early diagnosis of congenital Chagas' disease with a DOT assay using recombinant antigens with immunofluorescence antibody testing (IFAT) and an enzyme-linked immunosorbent assay (ELISA). The studies were performed using cord blood and sera of 12 infected newborns (group I) and 12 uninfected ones born to Trypanosoma cruzi-infected mothers (group II). Conventional IFAT and ELISA showed positive results for IgG at high titers, in infants and mothers of both groups; IgA antibodies were detected by ELISA in four of the infected infants and IgM was detected in two of them. All sera of the uninfected infants were negative for IgA and IgM in the ELISA. Application of a DOT assay using eight recombinant T. cruzi antigens allowed detection of specific IgA in the cord blood of six of the infected cases and IgM in eight of them. Repetition of these serologic tests in samples obtained during a monthly follow-up gave positive results for IgA in two of the initially negative infants of group I and for IgM in four of them. This means that diagnosis of congenital T. cruzi infection was confirmed, through demonstration of specific IgM, in all infected infants, and of IgA in eight of them. The importance of late detection of IgM in siblings born of infected mothers is discussed. The detection of IgM and IgA in sera obtained after birth is believed to be due to a congenital transmission of the parasite that occurred late in pregnancy. No IgA or IgM antibodies could be detected by the DOT assay in the sera of the negative controls. The IgA antibodies recognized predominantly shed acute-phase antigen, a marker of acute infection. Immune responses to IgM antibodies were produced against six of eight recombinant antigens. The results of the study indicate that the method used is highly specific and sensitive. It is concluded that the use of recombinant T. cruzi antigens greatly improves the early diagnosis of congenital Chagas' disease, especially through the detection of IgM. Although detection of IgM has been achieved in all of the cases in this study, additional study of IgA will be justified because some congenitally infected infants, who do not produce IgM, may develop specific IgA.