Plasma exchange (PE) with plasma infusion is the treatment of choice for thrombotic thrombocytopenic purpura (TTP) but doubts remain as to whether all kinds of plasma are equally effective. A multicentric cohort study was conducted to compare methylene blue-photoinactivated plasma (MBPIP) with quarantine fresh frozen plasma (qFFP) in the treatment of TTP. One hundred and two episodes of idiopathic TTP were included; MBPIP was used in 63 and qFFP in 39. The treatment schedule consisted of daily PE and costicosteroids, and the main end-point was remission status on day 8. Patients treated with MBPIP required more PEs (median: 11 vs. 5, P = 0·002) and a larger volume of plasma (median: 485 ml/kg vs. 216 ml/kg, P = 0·007) to achieve a remission, and presented more recrudescences while on PE therapy (29 of 63 vs. 8 of 39, P = 0·02) than those receiving qFFP. After adjustment for possible confounding factors, the use of MBPIP was associated with a lower likelihood of remission on day 8 [Odds ratio (OR): 0·17; 95% confidence interval (CI): 0·06–0·47] and a higher risk of recrudescence while on treatment (OR: 4·2; 95% CI: 1·6–10·8). In conclusion, MBPIP is less effective than qFFP in the treatment of TTP.