A series of bile acid derivatives were synthesized, purified and radiolabelled. These were [125I]cholylglycylhistamine [( 125I]CGH), [125I] cholyldiglycylhistamine [( 125I]CG2H), [125I]cholyltriglycylhistamine [( 125I]CG3H), and [125I]cholyltetraglycylhistamine [( 125I]CG4H). These derivatives were rapidly excreted unchanged into the bile of bile-fistula rats. In normal rats the 30-min cumulative excretion following intravenous administration was only 39.0 +/- 0.7% for [125I]CGH but greater than 80% for the three larger compounds. This marked difference in biliary recovery between CGH and the other larger compounds could be due to a threshold biliary permeability, and we postulated that the critical molecular weight threshold for effective biliary retention of such compounds falls between [125I]CGH (MW 683) and [125I]CG2H (MW 740). Increased permeability, involving a shift to a higher molecular weight threshold would then be anticipated to diminish biliary excretion of [125I]CG2H (MW 740) before exerting a major influence on the biliary excretion of [125I]CG4H (MW 854). We previously reported functional and morphological studies which suggest that ethinyl estradiol (EE) may alter the permeability of bile canalicular tight junctions. In this study we have looked for further evidence of a progressive permeability change in EE-induced cholestasis by observing the biliary excretion of CG2H and CG4H in rats. Treatment with EE (5 mg/kg/day) for 3 days (EE3) or with the injection vehicle propylene glycol for 7 days (C7) reduced biliary excretion to a significant extent when compared to 3-day controls (C3) but had no differential effect on the 30-min recoveries from bile of CG2H and CG4H, respectively: C3 (81.2 +/- 1.8% and 81.7 +/- 2.1%, P = CN): C7 (72.3 +/- 3.0% and 73.5 +/- 3.6%, P = NS): EE3 61.8 +/- 2.5% and 61.9 +/- 2.7%, P = NS). However, treatment with EE for 7 days significantly reduced the biliary recovery of CG2H (46.8 +/- 9%) compared to EE3 rats (P less than 0.0025) but there was no significant change of biliary CG4H recovery (61.0 +/- 2.5%, P = NS) compared with EE3 rats. These results are compatible with our hypothesis that EE-induced cholestasis is associated with a change of biliary permeability which, as it progresses, affects successively larger molecules.