Besides the accepted major risk factors for myocardial infarction (MI), cholesterol, hypertension and smoking, several other variables such as lipoproteins, apolipoproteins, fibrinogen and family history of MI, have been considered, but their usefulness as predictors of MI is controversially discussed. The Göttingen Risk Incidence and Prevalence Study (GRIPS) aimed to evaluate the independent impact of the latter in comparison to the established risk factors. GRIPS is a prospective cohort study, which included 5790 men, aged 40-59.9 years, without cardiovascular disease at baseline. Multivariate logistic regression models for the estimation of the MI risk based on the 10-year follow-up data from 97.4% of the study participants established LDL cholesterol as the strongest predictor of MI. It was followed by family history of MI, Lp(a), age, smoking, systolic blood pressure, HDL cholesterol (inversely related) and plasma glucose (P < 0.001). Apolipoprotein B as well as the ratios total/HDL cholesterol, LDL/HDL cholesterol and Apo B/AI were less effective predictors than LDL cholesterol and did not contribute independently to the estimation of MI risk. Similarly apoprotein AI was a weaker predictor of MI risk then HDL cholesterol. GRIPS is the first prospective cohort study which clearly justifies the key role of LDL cholesterol in preventive strategies. However, the data also give strong support for the additional consideration of other risk factors for a valid estimation of the MI risk for an individual subject.