The aim of this study is to assess melanoma risk in rheumatoid arthritis (RA). A literature review was performed, retrieving observational studies over 1990-2010 that provided estimates of relative risk of melanoma associated with RA, compared to the general population. We generated standardized incidence ratio (SIR) estimates across all studies, first pooling the data and then performing a random-effects model to generate the SIRs. We retrieved 713 citations; after reviewing the titles and abstracts, 124 were further reviewed, and of these, 11 met our inclusion criteria for analysis. Pooling the data, there were a total of 601 melanomas that were observed over 1,351,061 patient-years of follow-up, or 4.4 cases per 10,000 years. The expected number of melanomas over this interval was 596.2 for a pooled SIR of 1.01 [95% confidence interval (CI), 0.93, 1.09]. Excluding the two patient groups that were known to be exposed to biologics, the SIR estimate was unchanged (1.01; 95 CI, 0.93, 1.10). Our random-effects model similarly indicated an SIR for melanoma of 0.95 [95% credible interval (CrI), 0.86, 1.03] overall and 0.95 (95% CrI, 0.86, 1.04) excluding the two patient groups that were known to be exposed to biologics. These results do not highlight an important increased risk for melanoma in RA patients over-all, compared to the general population. It is not clear whether the risk is different for patients specifically exposed to biologic agents, although data are relatively few. Further study, especially of RA patients with a past history of melanoma, is warranted.