Direct HDL-C and LDL-C methods, which do not require the physical separation of lipoproteins, offer many practical advantages for routine clinical laboratory testing and are now widely utilized. Based on our previous work, we found, however, that they often yield results that substantially differ by the gold standard method, involving ultracentrifugation. In the past year, we have extended this work by assessing the impact of analytical errors in the calculation of Non-HDL-C in cardiovascular risk assessment. The results show that for patients with a dyslipidemia that Non-HDL-C, using a direct HDL-C assay, was superior to using a direct LDL-C in categorizing patients into cardiovascular risk categories. We also found that for some direct HDL-C assays that alternatives to the Friedewald equation for calculating LDL-C showed better correspondence to ultracentrifugation. We also completed a study and are now analyzing the data comparing apoB to LDL-P, as measured by NMR, in predicting cardiovascular events in the MESA population.