In order to evaluate the stability and reproducibility of syphilis serology qualitative results, the results reported on replicate specimens in the College of American Pathologists Quality Evaluation Program during 1978 were examined. There is evidence of good stability of syphilis serology specimens and reproducibility of qualitative results separated in time by as much as nine months for specimens that are nonreactive or highly reactive. However, there is evidence of lack of such stability and reproducibility of weakly reactive specimens. There is no striking difference in these results for the RPR card or VDRL test. The FTA-ABS test is more stable over time. There is some indication that the lack of stability and reproducibility of weakly reactive specimens is primarily a function of the stability of the specimen, with laboratory performance playing a secondary role.