Since an original report suggesting a protective effect of oral contraceptives on rheumatoid arthritis incidence, 12 studies from four countries have shown conflicting results. Using the method of DerSimonian and Laird all available studies (including two completed but not as yet published) were evaluated. Overall, the relative risk of oral contraceptive use on the development of RA were less than unity. When summary statistics for case-control studies were examined, a small protective effect was noted, but was not statistically significant.