Background: Profiling approaches in proteomics, such as surface-enhanced laser desorption/ionization (SELDI) mass spectrometry, are used in disease marker discovery. The aim of this study was to investigate the potential influence of selected preanalytical factors on the results obtained. Methods: Plasma samples anticoagulated with EDTA, citrate, or heparin, and serum samples from healthy volunteers were profiled by SELDI on CM10, immobilized metal affinity capture (IMAC) array with copper, and H50 chip surfaces. Using linear mixed-effects models, we examined the influence of elapsed time between venipuncture and sample separation (immediate to 24 h) and the type of serum tube used (Greiner Vacuette activator, gel serum separator, or plain tubes). We analyzed purified platelets, as well as platelet-poor and platelet-rich plasma samples treated with calcium and/or thrombin to determine the platelet contribution, directly or via the clotting process, to the profiles generated. We then used cluster analysis to identify samples with similar peak profiles. Results: Different plasma types and sera could be distinguished on the basis of cluster analyses of their spectral profiles. Elapsed time between venipuncture and separation of plasma and serum from blood samples altered the profiles obtained, particularly for serum