Severe enteritis and death consistently occur within 150 hours of a single intraperitoneal dose of 25 mg/kg methotrexate given to rats fed an elemental diet. Rats fed a regular chow diet show no clinical evidence of gastrointestinal morbidity or mortality at this dose. The etiology of this enhanced toxicity is not clear, however. It this study rats were randomized to receive an elemental diet (n = 45) or a regular chow diet (n = 45) for 7 days. They were then given a bolus dose of 3H methotrexate (25 mg/kg) intraperitoneally. At 20 and 40 minutes and at 1, 2, 12, 24, 48, 72, and 96 hours after injection, five animals from each group were anesthetized, and mean methotrexate levels in bile and serum, and tissue methotrexate concentrations in proximal, mid, and distal small intestine and in liver were obtained. Mean methotrexate levels in bile were significantly elevated in rats fed an elemental, chemically defined diet at 12, 24, 48, and 72 hours compared with regular diet fed rats. Mean serum methotrexate levels were elevated in both the inferior vena cava (IVC) and portal vein (PV) at 12 (IVC), 24 (PV), and 48 hours (IVC) in elemental, liquid diet fed rats. There were no statistical differences between methotrexate levels in rats fed an elemental diet or a regular chow diet in proximal or mid small intestine at any of the time points measured. In the distal small intestine, methotrexate levels were significantly elevated at 24 hours in rats fed an elemental, chemically defined liquid diet (P less than 0.03) compared with regular diet fed rats. Methotrexate levels in liver of rats fed an elemental, chemically defined liquid diet were significantly elevated at 24 and 48 (P less than 0.05) hours. Administration of an elemental, chemically defined liquid diet prolongs the enterohepatic circulation of methotrexate in rats and delays clearance of the drug from the systemic circulation. Thus, prolonged exposure of methotrexate to the cells of the intestinal mucosa explains the increased gastrointestinal side effects of the drug in rats fed an elemental, chemically defined liquid diet. If these results are applicable to man, elemental, chemically defined liquid diets are contraindicated as the sole nutritional source in patients receiving methotrexate. Such qualitative alteration of dietary intake may be a factor in the unpredictability of gastrointestinal side effects associated with high dose methotrexate administration.