Culture is currently considered the "gold standard" for detecting Chiamydia trachomatis infections. We evaluated the Syva MicroTrak enzyme immunoassay (EIA) and Gen-Probe PACE 2 tests, which detect chlamydial antigens and rRNA, respectively. These assays were compared with each other and with culture for the detection of C. trachomatis in cervical specimens obtained from 217 women attending a clinic for sexually transmitted diseases. The prevalence of infection was 22.1% by culture. The sensitivity, specificity, and positive and negative predictive values were 79.2, 98.2, 92.6, and 94.3%, respectively, for EIA. For PACE 2, the respective values were 77.1, 97.6, 90.1, and 93.7%. After corrections for two false-negative cultures, the sensitivities and specificities were 80 and 99.4%, respectively, for the EIA and 78 and 98.8%, respectively, for the probe assay. Quantitative evaluation of the results showed that false-negative results with either assay were associated with cultures that had low inclusion counts or were negative without subpassage. Analysis of nonculture results revealed that 2.3% of the EIA results and 4.6% of the probe assay results were within ±30%o of the respective assay cutoff values. These included four false-negative (one EIA and three probe) and two false-positive (one EIA and one probe) results. The Syva MicroTrak EIA and the Gen-Probe PACE 2 assay are comparable to but significantly less sensitive than culture. Use of a grey zone may help identify the need for repeat or confirmatory testing.