In an attempt to increase Chlamydiazyme (Abbott Laboratories) detection of Chlamydia trachomatis antigen and to establish the reproducibility of positive results, we carried out an investigation into the usefulness of testing duplicate specimens, of more aggressive endocervical specimen collection by using cytobrushes instead of swabs, and of the repeated testing of both specimens from patients with one or two positive results. Duplicate endocervical (female) and urethral (male) specimens, including one swab and one cytobrush specimen from 1,331 nonpregnant women, were collected from symptomatic and asymptomatic patients. Specimens were transported and tested for C. trachomatis antigen as specified by the manufacturer. Tests on all specimens from patients with positive results were repeated. Antigen was initially detected in one or both specimens from 210 (10.7%) of 1,968 patients, and repetition of the tests confirmed its presence in 198 (10.1%) of the patients, including all 183 patients in whom it was initially detected in both specimens. Initial results from at least 8 of the 12 patients with irreproducible antigen detection were most probably falsely positive. Results from 21 (10.6%) of the 198 patients for whom antigen detection was confirmed were repeatedly positive on only one specimen (9 [4.5%] on the second of the two specimens collected). Of 115 women from whom one swab and one cytobrush sample were taken and who had repeatedly positive results, antigen was detected in 7 (6.1%) only on the swab sample and in 4 (3.5%) only on the cytobrush sample. Use of the cytobrush does not appear justified with the Chlamydiazyme assay, and collection of duplicate specimens provided only a modest increase in detection of C. trachomatis. However, repeated testing of specimens when results from only one of two specimens are positive appears to be of clinical value.