The paper reports that high fluoride concentrations in the intestinal lumen hinders the absorption of this anion. This conclusion was verified with three different experimental models. Pharmacokinetic experiments done in human volunteers revealed that the bioavailability of fluoride from sodium fluoride (NaF, CAS 7681-49-4) enteric coated tablets was 33% of that of plain (immediate release) tablets. The latter findings were confirmed in rats receiving 1 ml of NaF solutions (40, 80 or 160 mmol/l) by gavage. The greatest AUC (area under the curve of fluoremia as a function of time) was obtained with an oral dose of 80 mumol of NaF. This parameter was significantly greater (p < 0.01) with 80 mumol than with 40 mumol NaF, but similar to that observed with 160 mumol. Fecal fluoride excretions (in the 24 h following a single dose of NaF) and the bone fluoride contents (found at the end of 30 days of treatment with 40, 80 or 160 mumol NaF/day), agreed with the AUC values. The rate of fluoride absorption (v, mumol/10 min) through the intestinal wall was investigated with perfused, isolated rat duodenum in vivo. Fluoride absorption increased between 0 and 10 mmol/l luminal fluoride and decreased with higher concentrations. Oxygen consumption of duodenal-tissue decreased exponentially between zero (1.12 mumol O2 min-1 g-1) and 10 mmol/l fluoride (0.45 mumol O2 min-1 g-1).