It has been reported that platelet expression and plasma levels of soluble P-selectin are increased in patients with unstable coronary artery syndromes. However, the origin of soluble P-selectin remains unknown. We sought to determine whether platelet expression of P-selectin correlates with plasma levels in the population of patients presenting to the emergency department with chest pain. In 338 patients presenting with chest pain to the emergency departments of three different hospitals, simultaneous soluble and platelet P-selection levels were determined using enzyme-linked immunosorbent assay (ELISA) and whole blood flow cytometry, respectively. Using regression analysis no correlation (R(2)=0.055) was found between soluble and platelet-bound P-selectin for the study population, including those patients with noncardiac chest pain (R(2)=0.019), unstable angina (R(2)=0.007), acute myocardial infraction (R(2)=0.033), congestive heart failure (R(2)=0.231), and gastrointestinal illness (R(2)=0.020). The platelet expression of P-selectin is unrelated to the level found in plasma in patients with acute chest pain, irrespective of the etiology of chest pain. Dissociation between platelet and soluble P-selectin suggests that the soluble form cannot serve as a surrogate marker to indicate platelet activation in the chest pain population.