Abstract Physicians are increasingly using nuclear magnetic resonance (NMR) analysis of lipoproteins to assess patients’ cardiovascular disease risk. For certain lipid measurements, like the standard lipid panel, which includes cholesterol, HDL-C, LDL-C, and triglycerides, there is debate about the necessity of fasting prior to a blood draw because the results may still be useful in the nonfasting state. Unfortunately, nonfasting specimens are poorly tolerated in NMR spectroscopy due to spectral overlap of chylomicrons and free fatty acids with other lipoprotein particles. However, it is unknown if valid lipoprotein levels can be extracted from these specimens. Approximately 5% of patient samples submitted for analysis have interferences that prevent analytical deconvolution of the NMR spectral data with the Numares Health AXINON LipoFIT assay. Retrospective review of patient samples from our laboratory showed that the interference rate increased with increasing triglyceride concentrations. Therefore, we hypothesized that sample centrifugation would reduce interference by low-density lipids on the NMR measurement. Preanalytical treatment consisted of centrifugation of 1 mL of serum at 13,000 rcf for 20 minutes followed by carefully pipetting 750 to 800 μL of serum from the lower half of the tube for the NMR measurement, thus avoiding the low-density chylomicrons and lipids at the top. To ensure that centrifugation does not disrupt the measured lipoproteins, a control set of 20 deidentified samples with previous NMR results received our preanalytical treatment and were reanalyzed. To test if centrifugation reduced interferences, a test set of 20 deidentified samples without NMR results due to interference were subjected to the preanalytical treatment and reanalyzed. The pre- and postspin results were analyzed, and the pre- and postspin spectra were compared. Our results showed that, for the control set, the average percent differences for all eight parameters determined by NMR (LDL, HDL, small-LDL, large-HDL, and large-VLDL particle numbers, and particle sizes of LDL, HDL, and VLDL) were within our acceptable limits of ≤10%. The pre- and postspin spectra of the controls were nearly identical. Fourteen of the 20 test samples (70%) achieved results after the preanalytical treatment. Visual inspection of test samples suggested that some, but not all, test samples had chylomicrons present due to the appearance of an upper white layer after centrifugation. The spectral comparison of pre- and postspin samples showed slight changes in the lipoprotein region. In conclusion, our study demonstrates that centrifugation is a promising preanalytical treatment to reduce the interference rate for NMR measurements of lipoproteins. Future goals include expanding this study to a larger sample set and determining if we can better identify samples that may benefit from this preanalytical treatment. In addition, the results suggest that there are other sources of interferences to investigate to improve the performance of the LipoFIT assay.