We have previously shown that sinusoidal reduced glutathione (GSH) efflux declines during development because of a declining maximum transport rate [Am. J. Physiol. 261 (Gastrointest. Liver Physiol. 24): G648-G656, 1991]. Because rat liver serves as the principal source of plasma GSH, we studied the response of plasma GSH to this declining inflow from liver. In immature (28- to 42-day) and mature (90- to 151-day) rats we injected tracer boluses of [35S]GSH intravenously and collected arterial samples over a 0.75- to 8-min interval while plasma GSH pool remained at steady state. Concentrations and radioactivities of GSH, oxidized glutathione (GSSG), cysteine (CYSH), cystine (CYSS), and cysteine-glutathione disulfides (CYSSG) and the radio-activities of proteins were measured in plasma. Our results show the following changes in plasma concentrations (microM): decreases in unbound (free) GSH (26.0 +/- 2.1 to 12.4 +/- 0.98; P < 0.001), total unbound GSH equivalents GSH + 2GSSG (29.1 +/- 2.1 to 15.3 +/- 1.2; P < 0.001), total reducible (unbound + bound) GSH (39.3 +/- 2.2 to 28.9 +/- 2.6; P < 0.025), and free CYSH (57.6 +/- 8.5 to 29.9 +/- 4.0; P < 0.05); no changes in GSSG (1.57 +/- 0.27 vs. 1.47 +/- 0.41), CYSS (36.7 +/- 12 vs. 43.4 +/- 17), and total unbound CYSH equivalents CYSH + 2CYSS (131 +/- 15 vs. 117 +/- 18); increases in total reducible (unbound + bound) CYSH (158 +/- 8.1 to 203 +/- 24; P < 0.05) and CYSSG (1.80 +/- 0.42 to 4.94 +/- 1.4 in microM GSH equivalents; P < 0.05). A concurrent decline occurred in irreversible disposal rate (IDR) of plasma GSH from 38.5 +/- 4.9 to 16.4 +/- 1.4 nmol.min-1.ml-1 (P < 0.001) as determined by compartmental analysis of tracer data. This 57% decrease in IDR parallels a decrease of 53% in the inflow of GSH estimated by perfused livers (17.0 to 8.0 nmol.min-1.ml plasma-1). However, perfused liver estimates do not match > 44-49% of plasma IDR. Thus perfused liver appears to underestimate the true rate of sinusoidal GSH efflux taking place in vivo. Some earlier arteriovenous data and our present portal vein-to-hepatic vein difference measurements appear to corroborate this view.