The steroid D-ring glucuronide conjugate estradiol-17 beta-D-glucuronide (E217G) but not the A-ring conjugate E23-glucuronide (E23G) has been shown to inhibit bile flow in the rat. To determine if primates are also sensitive to E217G-induced cholestasis, a noninvasive approach with the use of indocyanine green (ICG), a dye eliminated primarily by biliary excretion, was first validated in the rat and then applied to the rhesus monkey. ICG (16 mg/kg rat; 4 mg/kg monkey) was administered i.v. 10 min after an i.v. bolus dose of either E217G (5.5 or 11 mumol/kg), E23G (11 mumol/kg) or vehicle alone. In the rat, the elimination T1/2 of ICG was increased by the 11 mumol/kg dose of E217G (P less than .025), whereas E23G produced no significant change from vehicle control values. In the monkey, the 5.5 and 11 mumol/kg doses of E217G increased the T1/2 of ICG in a dose-related manner (P less than .005), whereas E23G was without effect. Plasma levels of total radioactivity demonstrated dose-dependent kinetics after the administration of a tracer dose and 11 mumol/kg of [3H]E217G. A rebound of plasma radioactivity was seen at 11 mumol/kg of [3H] E217G, the time course of which mimicked the time course of E217G-induced cholestasis. High-performance liquid chromatographic analysis of rat bile and plasma after the administration of [3H]E217G revealed primarily E217G and estradiol-3-sulfate-17 beta-D-glucuronide together with small amounts of three unidentified metabolites. In the monkey, only E217G and estradiol-3-sulfate-17 beta-D-glucuronide were observed in the plasma after the administration of [3H]E217G. In contrast to E217G, estradiol-3-sulfate-17 beta-D-glucuronide was choleretic in the bile duct-cannulated rat model. These data indicate that E217G is hepatotoxic in both rodents and nonhuman primates.