To identify the temporal changes occurring during progression and regression of atherosclerosis in nonhuman primates, we have studied the physicochemical and histological characteristics of arterial wall lesions during a 30-mo progression period of diet-induced hypercholesterolemia and during a 12-mo period of regression. Three groups of cynomolgous monkeys (Macaca fascicularis) were studied. Control groups were fed a basal chow diet for 18, 24, and 30 mo and were compared with progression groups that were fed a high-cholesterol-containing diet for up to 30 mo. Regression groups were fed a high-cholesterol diet for 18 mo to induce atherosclerosis and then fed monkey chow for up to 12 mo. The progression group monkeys were killed at 6, 12, 18, 24, and 30 mo, and the regression animals were killed at 24 and 30 mo (i.e., after 6 and 12 mo of being fed a noncholesterol-containing chow diet). Histology and morphometry, physical microscopy for cholesterol monohydrate crystals, foam cell and droplet melting points and chemical composition studies were completed on a large number of individual arterial lesions. Control animals had very little cholesterol ester, rare foam cells, and no extracellular cholesterol ester droplets or cholesterol crystals. During progression, the arteries first increased cholesterol ester content to produce high melting (approximately 45 degrees C) foam cell-rich lesions essentially devoid of cholesterol crystals. With time, the number of cholesterol crystals increased so that by 30 mo large numbers were present. Foam cells decreased with time but their melting temperature remained high while that of extracellular droplets fell to approximately 38 degrees C. Between 18 and 30 mo necrosis appeared and worsened. After 6-mo regression, unexpected changes occurred in the lesions. Compared with 24-mo progression, the chemical composition showed a relative increase in free cholesterol, a decrease in cholesterol ester and microscopy revealed large numbers of cholesterol crystals. Concomitantly, foam cells decreased and the melting temperature of both intra- and extracellular cholesterol ester markedly decreased. After 12-mo regression cholesterol decreased, cholesterol crystals and necrosis diminished and collagen appeared increased. Thus, during progression there is initially an increase in the number of foam cells containing very high-melting intracellular cholesterol ester droplets. By 30 mo, cholesterol crystals and necrosis dominate and high-melting foam cells appear only at lesion margins, suggesting that the initial process continues at the lesion edge. The lower melting point of extracellular esters indicates a lipid composition different from intracellular droplets. Thus, the changes observed in these animals generally reflect those predicted for progression of human atherosclerosis. During the initial 6 mo of regression, necrosis remains, the number of foam cell decreases, and cholesterol ester content decreases; however the relative proportion of free cholesterol content increases, and large numbers of cholesterol content are formed. Thus, large and rapid decreases in serum cholesterol concentration to produce regression in fact may result in the precipitation of cholesterol monohydrate and an apparent worsening of the lesions. More prolonged regression (12-mo) tends to return the lipid composition of the artery wall towards normal, partially reduces cholesterol crystals, and results in an improved but scarred intima.