The effect of the dithiols British Anti-Kewisite (BAL), dimercaptopropanesulfonic acid (DMPS), dimercaptosuccinic acid (DMSA) and a new metal binding agent 2,3-bis-(acetylthio)- propanesulfonamide (BAPSA) on the biliary excretion of arsenic in perfused livers of guinea pigs after acute experimental poisoning with As2O3 was investigated. Guinea pigs received As2O3, 10.0 mg/kg subcutaneously at 9 a.m. as a single injection. One hour after the injection the livers were perfused (2.5 ml x min.-1 x g-1 liver) with Krebs-Henseleit buffer and glucose for 80 min. After 40 min. of saline perfusion (control) 0.1 or 0.7 mmol/l BAL, DMSA, DMPS, or BAPSA were added to the perfusate and arsenic elimination in the bile and effluent perfusate was measured. The biliary excretion of arsenic in control livers between 40 and 80 min. was 0.7% of the total arsenic liver content before perfusion (= arsenic liver content after perfusion + portion excreted in the bile+perfusate). After antidote addition (0.1 mmol/l) the excretion was 0.2% for livers perfused with BAL, 6.8% for DMSA, 10.6% for DMPS, and 11.1% for BAPSA, respectively. After 0.7 mmol/l antidote the excretion of arsenic was 0.1% in livers perfused with BAL, 9.6% for DMSA, 12.3% for DMPS, and 13.3% for BAPSA, respectively. Except BAL, all compounds and most effectively BAPSA increased biliary excretion of arsenic. This indicates that excretion of arsenic which normally is mainly renal is shifted towards faecal excretion by the dithiols.