It has been suggested that thromboelastography TEG) can help in limiting or directing the appropriate use of blood products during surgery. However, the contribution of platelets to the TEG profile has not been studied in detail. Blood was taken from eight healthy subjects and eight patients with peripheral arterial disease (PAD). Immunomagnetic separation was achieved by the addition of Dynabeads labeled with a CD41 murine antibody (to the GPIIb/IIIa receptor) to achieve 90-100% depletion of platelets from blood. This was then titrated with whole blood to achieve platelet counts of approximately 0, 25 and 50% of the original count to compare with whole blood using TEG. Platelet function was also assessed by spontaneous platelet aggregometry (SPA) at baseline and at the 50% dilution. SPA, maximum amplitude (MA) and K time were significantly different in PAD patients compared to controls (P < 0.05). In both controls and PAD patients there was a strong linear correlation between Log10 [platelet count] and MA (r = 0.97 for controls, r = 0.89 for PAD) and K time (r = -0.86 for controls, r = -0.68 for PAD). Correlation between Log10 [platelet count] and R time was poor in both groups. The MA and K TEG parameters may be most useful for assessing platelet transfusion requirements.