OBJECTIVE To study: (a) the chlamydial antibody response (to the D-K serovars) using the micro-immunofluorescence (micro-IF) test in the following groups: (I) chlamydial genital infection only, (II) chlamydial ocular infection only, (III) combined chlamydial ocular and genital infection (oculo-genital infection), (IV) chlamydial ocular infection with chlamydia-negative non-gonococcal urethritis, (V) adenovirus conjunctivitis (control group 1), (VI) male partners of group I-IV with no chlamydial oculogenital infection or non-gonococcal urethritis (control group 2) (b) the cross reactivity of antibodies in patients' sera between the three chlamydial species and within the serovars of C trachomatis in those with culture-positive chlamydial oculo-genital infection. METHODS oculogenital (diagnostic) clinic at Moorfields Eye Hospital, London, UK. METHODS 209 consecutive patients attending the clinic with Chlamydia trachomatis oculogenital infection and 86 patients with adenovirus conjunctivitis (control group 1) and 55 male partners with no evidence of chlamydial oculogenital infection or non-gonococcal urethritis (control group 2). RESULTS Of all the patients with proven chlamydial oculogenital infection, 10.5% (22/209) and 94% (197/209) had IgM and IgG antibodies respectively. The geometric mean IgG antibody titres (GMT) were 1:98, 1:123, 1:245 and 1:101 in groups I to IV respectively. The IgG GMT values seen in control groups 1 and 2 were 1:45 and 1:36 respectively. Only 2/86(2%) patients in group V (control group 1) had IgG chlamydial antibodies of 1:32 and 1:64, whilst only 1/55(1.8%) and 4/55(7.3%) of patients in group VI(control group 2) had chlamydial IgG antibody titres of > or = 1:256 and > or = 1:128 respectively. A four-fold rise or fall in IgG antibody titre occurred in 56%(107/192) of patient groups I-IV over 2-6 weeks. Low titre cross-reactive antibody responses against different chlamydial species and serovars were commonly seen; 71%(148/209) of all patients showed cross-reactivity with Chlamydia pneumoniae or psittaci species or both, whilst 92% (193/209) of patients showed some level of cross reactivity to other pooled serovars of C trachomatis (A-C and L 1-3). CONCLUSIONS Serological diagnosis of chlamydial infection as evidenced by a positive IgM antibody response, high IgG titre (> or = 1:256) or > or = 4-fold rise or fall in IgG antibody titre was seen in 78%(163/209) of patients with culture-positive chlamydial oculogenital infection. Chlamydial IgG antibody titres of > or = 1:256 had a sensitivity of 42.6%, specificity of 98.2%, positive predictive value of 98.8% and a negative predictive value of 31% for chlamydial infection at any site, when considering groups I-IV and control group 2. In this study of 216 patients with conjunctivitis, a positive IgG antibody response (titre > or = 1:16) had a sensitivity of 98.5%, specificity of 97.7%, positive predictive value of 98.5% and a negative predictive value of 97.7%, for chlamydial conjunctivitis. Patients with dual chlamydial infection of conjunctiva and genital tract had a higher IgG GMT titre than those with ocular or genital infection alone: infection at a second site may produce an anamnestic response. Although the micro-IF test is a useful adjunct for the diagnosis of chlamydial infection, cross-reactivity between different chlamydial species and serovars is common. Chlamydial seroepidemiological studies should be interpreted with caution, as studies may attribute a serological response to a particular species or serovar in a setting where two or more are prevalent.