The alterations in lipid metabolism that occur with the use of oral contraceptives (OCs) have aroused considerable concern that OCs might increase the risk of premature atherosclerosis. However, most studies examining the role of OCs in atherogenesis were performed using earlier-generation preparations employing larger doses of sex hormones than present formulation. Therefore, we undertook a comparative and standardized determination of the effects on lipid metabolism of six modern, low-dose OCs. This open, randomized, comparative study included patients recruited at 21 study centers throughout Europe. Four hundred sixty-six women, aged 18-38 years, participated. They were randomly assigned to the following OC formulations:(1) norgestimate 250 micrograms + ethinyl estradiol (EE) 35 micrograms (Cilest); (2) norgestimate 180/215/250 micrograms + EE 35 micrograms (Tricilest); (3) desogestrel 150 micrograms + EE 20 micrograms = (Marvelon); (4) desogestrel 150 micrograms + EE 30 micrograms (Mercilon); (5) gestodene 75 micrograms + EE 30 micrograms (Femovan); and (6) gestodene 50/70/100 micrograms + EE 30/40/30 micrograms (Trifemovan). There were three parallel studies with six parallel patient groups. Fasting blood samples were drawn at baseline (between days 24 and 28) and on days 18-22 of cycle 6, and cycle 12. Sample were analyzed for total cholesterol,high-density lipoprotein (HDL) cholesterol, HDL2 cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides, apolipoprotein (apo)A1, and apoB at one central laboratory. Two hundred eighty-two women completed all 12 cycles and were included in the final evaluation. As expected, triglyceride and total cholesterol concentrations increased in all study groups but to lesser levels with the formulations containing gestodene. All OCs, except the monophasic gestodene preparation, slightly but significantly increased HDL. The HDL2 subfraction did ot change significantly except in the group using the monophasic gestodene preparation; in this group, the HDL2 subfraction slightly but significantly decreased. The LDL concentration increased slightly with the monophasic and triphasic norgestimate preparations and with desogestrel + 20 micrograms EE. The LDL/HDL ratio did not change significantly except with the use of the triphasic norgestimate preparation, in which case it decreased slightly. ApoA1 and apoB levels increased only slightly with all formulations. Importantly, while all of the OCs tested altered lipid levels to some extent, after 12 cycles there were no statistically significant differences in lipid effects among OC preparations. Modern, combined OCs that contain norgestimate, desogestrel, or gestodene all have some impact on lipid levels. However, it would appear likely that they do not contribute to atherogenesis in healthy women.