Study objectives: To assess the incidence of thrombocytopenia in surgical ICU patients, the factors associated with thrombocytopenia, the outcome of thrombocytopenic patients, and the possible mechanisms involved. Design: Prospective study. Setting: An 8-bed surgical ICU in an 885-bed teaching hospital, Patients: 147 consecutive patients admitted to the surgical ICU during a 6-month period, Main outcome measures: Incidence of thrombocytopenia (defined by a platelet count < 100,000/mm(3)), risk factors for thrombocytopenia, or death in thrombocytopenic patients identified by a stepwise logistic regression analysis, as well as the mechanisms involved, Results: Thrombocytopenia occurred in 52 patients (35%) with a mortality rate of 38%, compared with a 20%mortality rate in nonthrombocytopenic patients (p = 0.02). Sepsis, episodes of bleeding or transfusions, and an acute physiology and chronic health evaluation (APACHE) II score of > 15 were the independent risk factors identified for thrombocytopenia, The correction of thrombocytopenia mas a protective fatter I educing the risk of mortality in thrombocytopenic patients. Disseminated intravascular coagulation was found in 40% of thrombocytopenic patients, elevated platelet-associated IgG in 33%, and hemophagocytic histiocytes in 67%. Combinations of two of these mechanisms were demonstrated in one quarter of thrombocytopenic patients. Conclusions: Sepsis was the major independent risk factor identified, Thrombocytopenic patients had a higher ICU mortality due to the severity of overall clinical status. Bone man-ow examination could be diagnostic when no obvious causes are demonstrated. Thrombocytopenia probably reflects the severity and course of an underlying pathologic condition, as its correction appears to be a good prognostic factor.