Background : Clinical chemistry testing is influenced by a variety of preanalytical variables, including sample preparation. The presence of a diluted plasma layer at the top of primary tubes containing plasma citrate has recently been reported. However, no indication is available so far on the potential non-homogeneous distribution of clinical chemistry analytes during centrifugation of primary tubes containing lithium-heparin as an additive. Methods : A total of 40 lithium-heparin plasma samples were collected from volunteers and immediately centrifuged. An aliquot was obtained from the upper 0.4 mL of plasma (upper aliquot), 1.0 mL of plasma was discarded, and a second aliquot (lower aliquot) was obtained from the remaining plasma. The concentrations of alanine aminotransferase, albumin, alkaline phosphatase, amylase, amylase pancreatic, aspartate aminotransferase, direct bilirubin, total bilirubin, blood urea nitrogen, calcium, chloride, cholesterol, C-reactive protein (CRP), creatinine, creatine kinase, γ-glutamyltransferase (GGT), glucose, high-density lipoprotein-cholesterol, iron, lactate dehydrogenase (LDH), magnesium, phosphate, potassium, total protein, sodium, triglycerides and uric acid were assayed on a Roche/Hitachi Modular System P according to the manufacturer's specifications and using proprietary reagents. Sodium, chloride and potassium were measured on a Roche/Hitachi Modular System using indirect ion-selective electrode methods. Results : We observed a statistically significant difference between the upper and lower aliquots for CRP (3.88±0.67 vs. 3.94±0.68 mg/L; p=0.025), GGT (32.1±8.0 vs. 31.8±8.0 U/L; p=0.013), LDH (395±19 vs. 386±20 U/L; p=0.010) and triglycerides (1.29±0.09 vs. 1.27±0.09 mmol/L; p=0.001); results for the other analytes were not significantly different. In no case did the mean percentage bias recorded between aliquots exceed the current analytical quality specifications for desirable bias. Conclusions : The results of our investigation show that plasma layer stratification might occur in primary lithium-heparin tubes for a limited number of routine clinical chemistry tests, introducing a statistically significant bias in the measurement of GGT, LDH, triglycerides and CRP in the upper vs. the bottom section. When delayed testing is necessary for these parameters, we suggest that plasma should be separated after centrifugation and appropriately mixed before delayed/repeated analysis or aliquoting. Clin Chem Lab Med 2008;46:1180–2.