Serum samples were drawn from 14 healthy volunteers after an overnight fast and again after a meal. Each sample was divided into two aliquots. One aliquot was centrifuged at 20,000 rpm for 20 min, and a portion from the middle of the tube was removed by puncturing the tube's side. All samples were then added to cuvettes containing 0.14 M NaCl alone or 1, 2, 3, or 4% polyethylene glycol in 0.15 M KCl. The light scattering of each tube was measured. The centrifuged samples scattered 58% less light than the uncentrifuged controls. The range of values obtained with the centrifuged samples was also smaller as reflected in the standard deviation of the means obtained at each polyethylene glycol concentration. The coefficient of variation for replicate samples ranged from 0 to 16.6% (mean 5.8%), even when samples were processed differently. Storage at -70 degrees C significantly increased light scattering compared to baseline values. Serum samples were drawn from 44 patients with diseases that have been associated with immune complexes. Nephelometry selectively identified some groups of patients as abnormal. We conclude that preparative ultracentrifugation removes much of the background light scattering of serum samples, including that due to lipid, under conditions that do not sediment out immune complexes. Because it measures a physical property of immune complexes, this assay may provide information that is not available from biological receptor-based assays.