Serum samples from Bulgarian patients with physician-diagnosed erythema migrans (EM) (n=105) were examined using Borrelia burgdorferi ELISA (Boehring, Germany) after previous absorption with Treponema phagedenis. For IgM antibody detection sera were additionally pretreated with anti-IgG serum (RF absorbent). Serum samples of 93% of persons from healthy control group were IgM negative and all were IgG negative. Out of 105 patients with EM, 49% were IgM positive and 14 % were borderline. IgG ELISA showed positive results for 17% and borderline for 6% of the patients. Positive and borderline serum samples were examined further by immunofluorescent assay (IFA) and immunoblot test with recombinant B. burgdorferi proteins from strain PKo (B. afzelii) - p100, flagellin, OspA and OspC, and internal flagellin fragments from strains PKo and PBi (B. garinii) [B.Wilske, V.Fingerle, P. Herzer et al. 1993. Med. Microbiol. Immunol. 182:255]. IFA detected IgM antibodies against B. burgdorferi in 47 % of the positive and in none of the borderline by IgM ELISA serum samples as well as IgG antibodies in 83% of the positive and in 50% of the borderline by IgG ELISA samples. Presence of specific antibodies was confirmed by immunoblot in 71 % of the IgM ELISA postive and in 67 % of the IgG ELISA positive sera. In addition, anti-B. burgdorferi antibodies were detected in 60 % of the borderline by IgM ELISA serum samples. IgM serum reactivity was directed mainly against OspC antigen and flagellin and IgG antibodies were directed mainly against flagellin and p100. These findings clearly showed advantages of the ELISA test based on previous pretreatment of sera and capable to detect specific antibodies in more than half of patients with early Lyme borreliosis despite the well-known delayed immune response. IFA was less sensitive than ELISA in detection of anti-B. burgdorferi antibodies. An additional examination of ELISA borderline sera by immunoblot revealed more positive results. Serum reactivity to a single OspC antigen seems to be a sufficient criterion for positive IgM immunoblot.