OBJECTIVE: To determine if children with juvenile rheumatoid arthritis (JRA) are less likely to have been breast fed than controls. METHODS: Case-control study of data obtained from a survey of mothers 54 children with JRA and 79 playmates regarding breast feeding. Duration of breast feeding was tabulated and odds ratios (OR) with 95% confidence intervals (CI) were determined. RESULTS: OR for breast feeding in children with JRA was 0.40 (0.20-0.81, 95% CI) compared to playmates. For pauciarticular JRA (N = 28) OR was 0.31 (0.10-0.93); in polyarticular JRA (N = 24) OR was 0.60 (0.21-1.70). Lower OR for increased durations of breast feeding were noted in children with JRA. CONCLUSION: Children who have had JRA, especially pauciarticular JRA, are less likely to have been breast fed than controls, suggesting that breast feeding may have a protective effect on the development of JRA.