Nucleic acid amplification tests such as the BDProbeTec ET (BDPT) system are more prone to reproducibility problems than are antigen detection tests and culture. A repeat testing algorithm for all samples with method other than acceleration (MOTA) scores greater than or equal to the cutoff value (2000) was developed for the BDPT system and applied in a clinical laboratory setting. All positive samples were retested, and if the result of the second test was below the cutoff value, a third test was performed to resolve the discrepancy. Overall, 11 (5.3%) of 207 samples initially positive for Chlamydia trachomatis and 11 (10.7%) of 103 samples initially positive for Neisseria gonorrhoeae were not confirmed by repeat testing of the original sample. Poor reproducibility was associated with low-positive MOTA scores (2000 to 9999) for both analytes. Only 21 (80.8%) of 26 low-positive samples in the C. trachomatis test and 4 (33.3%) of 12 low-positive samples in the N. gonorrhoeae test retested as positive. The reproducibility of both tests with samples with initial MOTA scores of >or=10000 increased to 96.7%. The data suggest that retesting of low-positive samples is warranted and could reduce the number of potentially false-positive test results.