Metformin is eliminated by the kidneys, and metformin accumulation has always been noticed in oligo-anuric patients. We have reported an exception to the rule with the case of a metformin-treated patient having metformin accumulation contrasting with a mild increase in serum creatinine in the context of a volvulus of the sigmoid colon. This case prompted us to examine the association between intestinal occlusion and plasma metformin concentrations. For this purpose, we developed an experimental animal model of mechanical obstruction of the intestine. Rats were pre-treated during 3 weeks via drinking solution at a dose of approximately 100 mg/kg/day of metformin. They underwent at day 0 either sham-operation (n=7) or operation (n=8) to place a plastic tube around the ileum near the ileocaecal valve. Metformin administration was pursued on days 1, 2, and 3 giving a single dose of 100 mg/kg by intragastric gavage. Four days after the surgery, i.e. 24 h after the last metformin administration, the surviving intestinal obstructed rats (n=8) developed overt intestinal dilation but no biochemical abnormality compared to sham-operated animals (n=7; arterial lactate concentrations respectively 4.87 +/- 0.63 mmol/l and 3.97 +/- 0.30 mmol/l, NS, and serum creatinine concentrations 69.0 +/- 1.7 micromol/l and 68.7 +/- 1.9 micromol/l, NS). By contrast, there was a striking difference with regard to metformin concentrations, decreasing from 2.95 +/- 0.94 mg/l at day 0 to 0.12 +/- 0.03 mg/l at day 4 (p<0.001) in the sham-operated group but remaining unchanged (1.65 +/- 0.76 mg/l and 1.61 +/- 0.51 mg/l) in the operation group. In conclusion, this is the first experiment showing that intestinal occlusion may be responsible for metformin retention in the absence of renal failure. Whether this observation may be relevant to other drugs remains to be established.