BACKGROUND: Elevated level of low density lipoprotein-cholesterol (LDL-C) is one of the major risk factors for the development of coronary heart disease. Direct LDL-C determination method by immunoseparation (DLDL-C) recently developed is claimed not to be influenced by food ingestion. We re-evaluated the effects of diet and storage conditions for this method. METHODS: Samples were collected from thirty-two medical college students before and after meal to study the effects of diet on this method. We compared the difference of LDL-C of filtered samples between refrigerated and frozen state. We also compared direct and indirect calculated measurements of LDL-C with ultracentrifugal beta-quantification (BQLDL-C) method. RESULTS: Morning 2-hour-postprandial specimen can be acceptable with no minimal significant bias, but afternoon 2-hour or 4-hour-postprandial specimen cannot be recommended due to significant negative bias (8.6-9.6%). Storage of filtered samples showed no significant difference between frozen and refrigerated state. Calculated LDL-C when triglyceride level is more than 400 mg/dL was not reliable due to large proportional and constant bias. In contrast, DLDL-C showed good accuracy comparing with BQLDL-C (y=0.909x+3.3, r=0.869, n=9, x=BQLDL-C, y=DLDL-C). CONCLUSION: In conclusion, morning two-hour postprandial specimens can be acceptable for DLDL-C, but afternoon postprandial specimens may not be recommended due to significant negative bias. DLDL-C seems to be reliable and useful especially for hypertriglyceridemic patients or follow-up cases of hypercholesterolemia with normal triglyceride or HDL-C levels.