The study assesses the clinical features of individuals that best enable an effective selective screening of the adult population for dyslipidemia. The results of the population-based 1990 Saskatchewan Heart Health Survey were examined. Dyslipidemia was defined as a total cholesterol (TC) to high-density lipoprotein cholesterol (HDL) ratio of > or = 5. In total, 805 men and 782 women, 18-74 years of age, had their plasma cholesterol measured. Using TC screening of the entire population as recommended by the Canadian Consensus Conference on Cholesterol would correctly identify 79% of those with dyslipidemia (sensitivity) and 67% of those with a normal profile (specificity). However, if one performs lipoprotein analysis on only those with a high waist-to-hip circumference ratio (WHR), 44% of the population would need to be screened to correctly identify 71% of those with dyslipidemia (sensitivity) and 66% of those with a normal profile (specificity). A high WHR is an especially strong predictor of dyslipidemia in young adults. In conclusion, a high WHR may provide an effective means of selective screening of the adult population for dyslipidemia.