Zoster immune plasma (ZIP) was evaluated for treatment of cutaneous disseminated zoster in immunocompromised hosts. Twenty patients were studied: 13 were enrolled in a double-blind protocol, five received ZIP under an open protocol, and two were observed without receiving a transfusion. In the double-blind study, eight patients actually received ZIP; five were given plasma lacking varicella-zoster virus antibodies (control plasma). The clinical course of zoster in the group given ZIP was the same as that of patients given control plasma or no transfusions. Because ZIP did not alter the clinical course of zoster and because zoster patients produced high-antibody titers without ZIP, we concluded that ZIP is not useful for treatment of cutaneous disseminated zoster and should be reserved for prevention or modification of varicella in exposed, susceptible immunocompromised patients.