Retinyl ester hydrolase activity was studied in 25 7-wk-old rats that had been previously fed no vitamin A for 1 wk and then were fed 0, 5, 24, 60 or 240 micrograms retinol/d for 2 wk. This treatment produced rats with vitamin A reserves from depleted to normal (0.03 to 184 micrograms/g) and serum retinol concentrations of 4-67 micrograms/dl. Hydrolase activity in liver homogenates was optimal when assayed with 275 mM CHAPS (3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate), 2 mg/ml Triton X-100 and 5 mM retinyl palmitate at pH 7.0. By use of these concentrations and conditions, retinyl ester hydrolase activity displayed relatively low interanimal variability (fivefold). Furthermore, activity was independent of serum or liver vitamin A concentrations. When CHAPS in the assay was replaced with equimolar sodium cholate, or with reduced concentrations of CHAPS and retinyl palmitate, retinyl ester hydrolase activity was more variable, i.e., 37- and 23-fold, respectively. Furthermore, in the latter case, hydrolase activity was threefold higher in rats with liver vitamin A reserves less than 10 micrograms/g. Thus, the type and concentration of bile salt used to assay retinyl ester hydrolase affect its interanimal variability. Furthermore, hydrolase activity measured in reduced concentrations of substrate and detergent is elevated in rats with low liver reserves of vitamin A.