BACKGROUND: Reports of deterioration and death after platelet (PLT) transfusions in patients with thrombotic thrombocytopenic purpura (TTP) have led to recommendations that they should not be given except for life-threatening hemorrhage. STUDY DESIGN AND METHODS: Published reports of PLT transfusions in patients with TTP were systematically reviewed and data from the Oklahoma TTP-HUS Registry, an inception cohort of 382 consecutive patients, 1989 through 2007, were analyzed. RESULTS: A systematic review identified 34 publications describing outcomes of patients with TTP after PLT transfusions: 9 articles attributed complications to PLT transfusions, 4 suggested that they may be safe, and 21 articles did not comment about a relation between PLT transfusions and outcomes. Fifty-four consecutive patients from the Oklahoma TTP-HUS Registry were prospectively analyzed. ADAMTS13 activity was less than 10 percent in 47 patients; also included were 7 patients whose activity was not measured but who may have been deficient. Thirty-three (61%) patients received PLT transfusions. The frequency of death was not different between the two groups (p = 0.971): 8 (24%) patients who received PLT transfusions died (thrombosis, 5; hemorrhage, 1; sepsis, 2) and 5 (24%) patients who did not receive PLT transfusions died (thrombosis, 4; hemorrhage, 1). The frequency of severe neurologic events was also not different (p = 0.190): 17 (52%) patients who received PLT transfusions (in 5 of these 17 patients, neurologic events only occurred before PLT transfusions) and 7 (33%) patients who did not receive PLT transfusions. CONCLUSION: Evidence for harm from PLT transfusions in patients with TTP is uncertain.