The major high density lipoprotein (HDL) apolipoprotein, apoA-I, was knocked out by gene targeting in ES cells to provide a model for the study of HDL metabolism and its relationship to plasma and tissue cholesterol metabolism. HDL and non-HDL cholesterol (HDL-C) were reduced in apoA-I-deficient mice. Feeding a high fat-high cholesterol diet raised HDL-C minimally in apoA-I knockout compared to the large increase seen in control mice, suggesting an interaction between diet and apoA-I genotype. In apoA-I-deficient mice, HDL was normal in size but altered in composition. Compared to control mice there was more triglyceride and free cholesterol and less cholesteryl ester (CE), suggesting that apoA-I-deficient HDL is a poor substrate for hepatic lipase and lecithin:cholesterol acyltransferase (LCAT). The metabolic basis of the low HDL-C levels in the apoA-I knockout mice was decreased flux into the HDL CE pool. The absolute delivery of HDL CE to both peripheral tissues and liver was also decreased. As tissue cholesterol levels and synthesis were unchanged, the decreased flux of cholesterol into the HDL CE pool was most likely due to decreased efflux of cholesterol from the peripheral tissues and decreased functional LCAT activity. The low HDL-C state in the apoA-I-deficient mouse was associated with an absolute decrease in unidirectional transport of cholesterol from peripheral tissues to the liver but this did not lead to cholesterol accumulation in the periphery or a cholesterol deficit in the liver; nor was there altered peripheral tissue HMG-CoA reductase activity. The only sign of decreased cholesterol flux to the liver was a 2.3-fold decrease in liver cholesterol 7 alpha-hydroxylase mRNA, suggesting decreased bile acid synthesis. In the apoA-I knockout mouse model it appears that low HDL levels create a new steady state in which decreased cholesterol is delivered to both peripheral tissues and the liver.