Isolated non blood-perfused intestinal segments from normal and iron-deficient rats were used in vitro. A modification of the luminal perfusion method according to Fisher and Parsons allowed the comparison of iron and transferrin quantities in the serosal fluid at 15 min intervals. Iron transfer in jejunal and ileal segments was directly proportional to the luminal iron concentration within a dose range of 1 to 100 mumol/l, did not show saturation characteristics and was linear over time. Jejunal segments from iron-deficient rats transferred about twice as much iron as the jejunal controls. In ileal segments there was no difference in iron transfer between iron-deficient and control rats; in both cases transfer amounted to approx. 10% of jejunal controls. An exponential correlation was found, when the decreasing transferrin content of the tissue was plotted against the cumulative water transport. Transferrin and albumin release from jejunal and ileal segments into the absorbate cumulated asymptotically, which is typical for wash-out phenomena. As iron transfer cumulated linearly while transferrin release cumulated in an asymptotic manner, the capacity of transferrin to bind iron ions is exceeded roughly 100 times by molar equivalents of iron in the last absorbate fractions. Independence of iron transfer from mucosal transferrin quantities is concluded. As the molar transferrin/albumin ratios do not show significant differences between plasma and the sequence of absorbate samples, a wash-out from the gut's interstitial space is assumed, which makes plasma the most likely origin of transferrin in the mucosa.