Depending on radionuclide characteristics, radioimmunotherapy (RIT) relies on radioactivity to destroy cells distant from immunotargeted cells. Therefore, even heterogeneous tumors (for antigen recognition) can be treated, because not all cells have to be targeted. Substantial complete response rates have been reported in patients with non-Hodgkin's lymphoma. Much more modest results have been reported for patients with bulky solid tumors, e.g. adenocarcinomas. The radiation doses delivered by targeting antibodies are generally too low to achieve major therapeutic responses. Dose escalation is limited by myelotoxicity, and higher doses need to be delivered to neoplasms less radiosensitive than lymphomas. Various trials for both systemic and regional RIT have been reported on. Intraperitoneal administration has been applied for colorectal and ovarian carcinomas. Our own results indicate that, e.g., intraperitoneal pseudomyxoma can be treated with RIT. Myelotoxicity can be reduced by anti-antibody-enhancement, 2- and 3-step strategies, bispecific monoclonal antibodies (MAbs), and extracorporeal immunoadsorption. The radionuclide has to be selected properly for each purpose; it can be a beta-emitter, e.g. I-131, Y-90, Re-188, Re-186, Lu-177 or Sm-153, an alpha-emitter At-211 or Bi-212 or an Auger-emitter, e.g. I-125, I-123. One major problem with RIT, besides slow penetration rate into tumor tissue and low tumor-to-normal tissue ratio, is the HAMA response, which can be partly avoided by the use of humanized MAbs and immunosuppression. However, RIT will be, because of all the recent developments, an important form of cancer management.