The analytical performance of the Cholestech L.D.X. lipid analyzer for total and HDL-cholesterol and triacylglycerols in heparinized venous blood was evaluated, using two Cholestech L.D.X. analysers and two reagent cassette lots. Within-day and day-to-day precision were checked with commercial quality control sera. Within-day coefficients of variation (CVs) were 2.0 to 4.7% for total cholesterol, 3.4 to 5.5% for HDL-cholesterol, and 2.1 to 4.8% for triacylglycerols. Between-day CVs were < or = 6.0% for triacylglycerols, and < or = 7.0% for HDL-cholesterol. A method comparison study, according to the NCCLS EP9-P guidelines, was performed for all three analytes. Cholestech whole blood values (y) were compared with serum values (x) generated by a standardized enzymatic method for cholesterol, and by current state-of-the-art methodology for HDL-cholesterol and triacylglycerols. Correlation coefficients were all > 0.98. Mean slopes and intercepts for the Passing & Bablok regression equations of total and HDL-cholesterol were not significantly different from one and zero, respectively. Overall, the Cholestech means differed by < 1.8% for total cholesterol, and < 5% for HDL-cholesterol versus the comparison method means. For total cholesterol the National Cholesterol Education Program (NCEP) requirements for accuracy and precision of < or = 3% were met. For triacylglycerols, significant negative intercepts ranging from -0.26 to -0.33 mmol/l were observed, with slopes significantly greater than one.