Results from three randomized placebo-controlled trials were combined in a meta-analysis to compare the clinical utility of four advanced drug therapy agents used to treat juvenile rheumatoid arthritis (JRA): D-penicillamine (10 mg/kg/d), hydroxychloroquine (6 mg/kg/d), auranofin (oral gold, 0.15 to 0.20 mg/kg/d), and two low dose levels of methotrexate [5MTX, 5 mg/M2/wk; 10MTX, 10 mg/M2/wk]. A total of 520 children with JRA were enrolled into these trials. Only 10MTX resulted in significantly greater improvement than placebo in variables that assess effectiveness: physician's global assessment, a composite index, and erythrocyte sedimentation rate. Treatment effect sizes were the largest in the 10MTX group for all articular disease indices. The short-term safety profiles were similar across all treatment groups. It is concluded that the current trend among pediatric rheumatologists to use oral methotrexate as the first advanced drug therapy in JRA is appropriate and that the minimum effective dose is 10 mg/M2/wk.