High density lipoproteins (HDL) play a key role in preventing atherosclerosis, the main pathogenesis for cardiovascular diseases (CVD) which impair the health of coronary heart, cerebrovascular system, and peripheral arteries and other tissues. HDL transport excess of cholesterol from vascular tissues in the process of reverse cholesterol transport (RCT) to the liver for secretion and disposal. As a group of particles with great heterogeneity, HDL have been defined in their structure, composition and function by numerous studies; however, further work is necessary to understand how different types of HDL particles function as acceptors for cellular cholesterol. To gain an insight into these properties of HDL particles, physically defined populations of plasma and reconstituted HDL particles were examined in this project. Our results demonstrated that individual subspecies of plasma HDL were very different in their capacity for exporting cellular cholesterol from cultured cells; small plasma HDL (3b, 3c) and reconstituted HDL (< 8nm) particles were more active than larger plasma and rHDL particles for ABCA1-mediated cholesterol efflux, while ABCG1-dependent cholesterol efflux was related to the amount of HDL phospholipid supplied, independent of particle size. We demonstrated that ABCA1 played a critical role in cellular cholesterol efflux from both human and mouse macrophages, whereas ABCG1 contributed to efflux only in mouse macrophages. We made a preliminarily comparison of the efflux capacity of apoB-depleted serum using mouse and human macrophage cell models. Finally, the roles of ABCA1 and ABCG1 in cholesterol export from a cancer cell line (HeLa) were studied by stable silencing of these transporters. We showed that ABCA1-dependent cholesterol efflux was very low in these cells and that knock-down of either ABCA1 or ABCG1 had little effect on cholesterol efflux to cholesterol acceptors. These observations contrast to those in macrophages which rely strongly on ABC transporters for reverse cholesterol transport. It suggests that RCT from HeLa cells may be controlled at a different level, largely independent of LXR (Liver X receptor) regulation.