Abstract Objective To summarize the diagnostic accuracy of self-reported osteoarthritis (OA), rheumatoid arthritis (RA), and arthritis (i.e., unspecified) in the general adult population. Study Design and Setting A systematic literature search identified studies reporting diagnostic data on self-reported diagnosis of OA, RA, or arthritis in adults in population-based or primary care samples. Index tests included any form of participant-reported presence of the condition. Reference tests included rheumatologist, physician, or health professional examination; medical record review; physician interview; laboratory tests; or radiography. Relevant articles were scored using the QUADAS tool. Diagnostic values were summarized using pooled estimates for sensitivity and specificity. Results The search strategy identified 16 articles: 11 for OA, 5 for RA, and 4 for arthritis. Four of 16 articles scored high on quality. The pooled sensitivity and specificity were 0.75 [95% confidence interval (CI): 0.56, 0.88] and 0.89 (95% CI: 0.77, 0.95) for OA, 0.88 (95% CI: 0.59, 0.97) and 0.93 (95% CI: 0.66, 0.99) for RA, and 0.71 (95% CI: 0.59, 0.80) and 0.79 (95% CI: 0.65, 0.89) for arthritis. There were not enough studies to conduct meta-analyses for joint-specific OA. Conclusion The accuracy of self-reported OA and RA is acceptable for large-scale studies in which rheumatologist examination is not feasible. More high-quality studies are required to confirm the accuracy of self-reported arthritis and joint-specific OA.