Yolk-sac inoculation of embryonated eggs was superseded 25 years ago by the use of cell cultures (often McCoy) for the isolation of Chlamydia trachomatis. Centrifugation of specimens onto the cell monolayers was shown to increase sensitivity, but little of late has further improved sensitivity which is at least ten-fold greater than that of eggs. However, culture is slow and labour intensive so that non-cultural techniques without these drawbacks have come to dominate. Direct fluorescent antibody (DFA) tests are rapid and have sensitivities that range from 70% to 100% for men and 68% to 100% for women, and specificities that range from 87% to 99% for men and 82% to 100% for women; if the tests are read by competent observers the values are at the top end of the ranges. The detection rate may be enhanced even further by relatively low-speed centrifugation of specimens before staining. Skilled reading is not a feature of enzyme immunoassays (EIAs) which according to the literature have sensitivities that range from 62% to 97% for men and 64% to 100% for women, and specificities that range from 92% to 100% for men and 89% to 100% for women. However, comparison against poor reference tests is responsible for most of the higher values and the clinician should not be misled into believing that EIAs have excellent sensitivity; the lower values in the ranges are closer to reality. Furthermore, EIAs that are being designed for use by general practitioners should be regarded with the greatest caution since lack of sensitivity means that chlamydia-positive patients will go undetected. The polymerase chain reaction (PCR) is not bedevilled by insensitivity but it is no more sensitive than the most sensitive cell culture or DFA tests. PCR is unsuitable for routine diagnosis but has a place as a research tool. For men, examination of "first-catch" urine samples by the best of the non-cultural procedures provides an acceptable non-invasive approach to diagnosis; for women, the value of examining urine may be less, but needs to be thoroughly tested. However, there is little doubt that a Cytobrush used to obtain cervical specimens holds no practical advantage over a swab. Serological tests are reliant on the provision of paired sera for making a diagnosis; high antibody titres in single sera may be suggestive of an aetiological association in deep-seated chlamydial infections (epididymitis, arthritis, salpingitis, etc), but unequivocal interpretation is unusual, particularly in an individual case, since the distinction between a current and past infection is problematical.