Cyclosporine A (CSA) and tacrolimus (FK-506) are both associated with thrombotic microangiopathy (TMA) in allogeneic BMT recipients, although it is not known which drug is more likely to cause the syndrome. The optimal treatment of BMT-associated TMA is also not known. To estimate the risks, predisposing factors, and outcomes of TMA, data were analyzed from two cohorts of BMT patients who had received CSA or FK-506 in our institution with the same clinical definition for TMA. TMA was diagnosed in 11 of 55 patients (CSA, 3 of 24; FK-506, 8 of 31). The daily risk of developing TMA was 0.12 percent for patients receiving CSA and 0.26 percent for those receiving FK-506 (p = 0.16, chi-square). Among patients receiving FK-506, sibling donor BMT recipients were as likely to develop TMA as matched unrelated donor recipients. Serum CSA and FK-506 concentrations were not elevated above the therapeutic range in most patients with TMA. The blood urea nitrogen to serum creatinine ratio was elevated in patients with TMA. Despite daily plasmapheresis, 9 of 11 patients died without resolution of TMA; however, the causes of death were multifactorial, including GVHD. Histologic evidence for TMA was absent in 1 patient who died with persistent clinical signs attributed to microangiopathy. In this study, a high incidence of TMA was found in patients receiving either CSA or FK-506 following BMT, with uniform diagnostic criteria and strict monitoring. Neither drug showed elevated levels in most patients with TMA. Plasmapheresis was unsuccessful in reversing most cases of TMA.