In the WHO MONICA Project (Monitoring of Trends and Determinants of Cardiovascular Disease) total cholesterol was measured in representative samples from 51 study populations in 26 countries. The biochemical measurements were done locally by the collaborating centres' laboratories. Differences in measurement procedures among the populations were found in the following factors: fasting status, posture of the subject, tourniquet use, use of serum or plasma, storage conditions, and the analytical method itself. This paper gives an overview of the methods used, and discusses the possible effects of the differences on the comparability of the results. The use of a posture other than that recommended and the use of EDTA (ethylene diaminetetraacetate) plasma are considered to be the most important factors, and were found in 9 out of the 51 populations.