A prospective clinical trial was conducted in three centres to assess the effects of the type and dose of progestogen, the dose of estrogen and the progestogen-to-estrogen ratio of oral contraceptives on lipid metabolism. The preparations selected contained levonorgestrel 250 micrograms + ethinyl estradiol 50 micrograms (Neogynon), levonorgestrel 250 micrograms + ethinyl estradiol 30 micrograms (Eugynon 30), levonorgestrel 150 micrograms + ethinyl estradiol 30 micrograms (Microgynon) or norethisterone acetate 1 mg + ethinyl estradiol 50 micrograms (Minovlar). Four-hundred-and-seven premenopausal women were randomly assigned to one of the four pill groups and compared to a control group of 119 users of a CuT220c intrauterine device. Total cholesterol, HDL-cholesterol, LDL-cholesterol and total triglycerides were monitored and the analysis includes the data of those who were followed over 48 weeks, 241 OC users and 87 IUD users. 250 micrograms of levonorgestrel were found to induce more unfavourable lipid changes in terms of atherosclerotic risk than 1mg of norethisterone acetate. Levonorgestrel was found to have a dose-effect on HDL-cholesterol and LDL-cholesterol serum levels, while ethinyl estradiol had a dose-effect on serum triglycerides. HDL-cholesterol was related to the progestogen-to-estrogen ratio. Most of these findings were consistent across centres. Finally, some comments are made on the implications of the study results on the design of future lipid studies.