The hypocholesterolemic potential of peroxisome proliferator-activated receptor (PPAR) pan-activator bezafibrate has been documented. However, in addition to uncertainty about the contribution of PPAR alpha to its effect, there is a marked discrepancy in bezafibrate dosages used in previous rodent experiments (> or = 50 mg/kg/day) and those in clinical use (< or = 10 mg/kg/day). To investigate the association between bezafibrate-induced cholesterol reduction and PPAR alpha activation, wild-type and Ppar a-null mice were treated with bezafibrate at high (100 mg/kg/day) or low (10 mg/kg/day) doses and analyzed. High-dose treatment decreased hepatic cholesterol content in wild-type mice, but increased serum cholesterol concentration. In liver samples, simultaneous increases in the expression of numerous proteins involved in cholesterol biosynthesis and catabolism, as well as cholesterol influx and efflux, were observed, which made interpretation of phenotype changes subtle. These complicated responses were believed to be associated with intensive PPAR activation and accompanying up-regulation of liver X receptor alpha, farnesoid X receptor, and sterol regulatory element-binding protein 2 (SREBP2). In contrast, low-dose bezafibrate treatment decreased serum and hepatic cholesterol concentrations in a PPAR alpha-independent manner, probably from suppression of SREBP2-regulated cholesterogenesis and enhancement of cholesterol catabolism due to elevated 7alpha-hydroxylase levels. Interestingly, the low-dose treatment did not affect the expression of PPAR target genes or number of peroxisomes, suggesting the absence of PPAR activation. These results demonstrate that the action of bezafibrate on cholesterol metabolism may vary with dosage, and that the cholesterol-reducing effect found in mice at dosages similar to those administered to humans is independent of significant PPAR activation.