Background :Serum high density llpoproteln (HDL)-cholesterol level IS used as an assessment of the risk of coronary heart disease. In this study, we evaluated direct measurement of HDL-cholesterol in serum with polyethylene-modlfled enzymes and sulfated a-cyclodextrin Method :We evaluated the precision, the lower limit of detection, the recovery rate, the linearity, the Interference for hemoglobin and the comparlslon with the result of HDL-cholesterol measured by selective preclpltatlon method We also studied the specificity of this direct method for very low density llpoproteln (VLDL) and low density lipoprotein (LDL) Result :The total ImprecIsIon was 3 8% (low), 3 5% (mlddle), 3 2% (high) The lower limit of detectjon was 0 mg/L The recovery rate was satisfactory The linearlty was also (r`=O 99) This method showed a good correlation (r`=O 97) with the selective preclpltatlon method in HDL-cholesterol measurement VLDL-cholesterol (up to 300 mg/L) Increased HDL-cholesterol only less than 3% but Increased VLDL-cholesterol to 400 mg/L, more than 750 mg/L caused 5% and 15% of overestlmatlon of HDL-cholesterol, respectively LDL-cholesterol (142-l ,073 mg/L) Increased or decreased HDL-cholesterol by some degree (about 15%). Hemoglobin (up to 3,000mg/L) did not influence this assay Conclusion :The drrect measurement of HDL-cholesterol IS satisfactory method in HDL-cholesterol measurement In good analytlcal performance and may be anticipated to reduce workload of laboratory because the sample pretreatment IS not necessary