OBJECTIVE To determine the bioavailability of oral methotrexate (MTX) in patients with juvenile rheumatoid arthritis in the fasting and fed states. METHODS Each patient randomly received their usual weekly MTX dose either orally (po) after an overnight fast, po immediately after a breakfast of their choice, or intravenously (iv) on 3 consecutive weeks. Blood samples were taken at 0, 0.5, 1, 1.5, 2, 3, 4, and 6 h after po and 0, 0.08, 0.25, 0.5, 1, 1.5, 2, 3, 4, and 6 h after iv administration. RESULTS Fourteen patients (10 female) aged 2.8 to 15.1 yrs completed the study; the results of 13 patients were evaluable. The mean elimination rate constant was 0.27 +/- 0.065, 0.26 +/- 0.067, and 0.25 +/- 0.11 h-1 after po fasting, po fed, and iv administration, respectively. The total area under the serum concentration vs time curve was 1.87 +/- 0.83, 1.50 +/- 0.51, and 1.85 +/- 0.80 mumol/l.h after po fasting, po fed, and iv administration, respectively. The maximum serum MTX concentration (Cmax) was 0.65 +/- 0.33 and 0.39 +/- 0.18 mumol/l after po fasting and po fed administration, respectively (p = 0.0022). The time to Cmax was 0.94 +/- 0.40 and 1.32 +/- 0.68 h after po fasting and po fed administration, respectively (p = 0.1464). The bioavailability of oral MTX while fasting was 1.1 +/- 0.51, while that after a meal was 0.88 +/- 0.35 (p = 0.0211). CONCLUSIONS These data indicate greater oral bioavailability of MTX in the fasting state. We recommend that children receive MTX on an empty stomach.