The biological variation of several relative lipid quantities, calculated as the ratios between the concentrations of various serum lipids and apolipoproteins, has been estimated over a one-year period. The medians of the within-subject biological coefficient of variation, separated by sex when significant differences exist, were 15.4% for [apolipoprotein A-I]/[apolipoprotein B], 6.8% for [high-density lipoprotein (HDL)-cholesterol]/[cholesterol], 10.5% and 17.6% (women and men, respectively) for [HDL2-cholesterol]/[HDL-cholesterol], 13.6% for [HDL2-cholesterol]/[HDL3-cholesterol], 10.6% for [low-density lipoprotein (LDL)-cholesterol]/[apolipoprotein B], 10.6% and 8.7% (women and men, respectively) for [LDL-cholesterol]/[cholesterol], and 6.3% for [LDL-cholesterol]/[HDL-cholesterol]. From these data, we have calculated the critical difference for significant change detection, the index of individuality, and the goal for the between-day imprecision. Concerning within-subject biological variation, the best ratios for the detection of risk of coronary heart disease and the monitoring of intervention are [LDL-cholesterol]/[HDL-cholesterol] and [HDL-cholesterol]/[cholesterol]. The index of individuality obtained in this study indicates that the use of population-based reference values is inadequate for interpreting the ratios studied.