This article reviews the current status, including problems and some proposed solutions, of the targeting of cancer with radioactive antibodies for use in antibody imaging (radioimmunodetection) and radioimmunotherapy. The problems and results are similar for purified polyclonal and murine monoclonal antibodies. Foremost among the problems are the low accretion of antibody (0.01% to 0.001% of injected dose per gram) in tumor and the nonspecific deposition of radiometals that are more ideal for imaging or therapy after antibody conjugation. Despite these limitations, radioimmunodetection appears to be a safe and useful method, even at this early stage of development. Sensitivity, specificity, and accuracy rates of 80% to 90% have been achieved in some studies involving radioiodine labels and polyclonal or monoclonal antibodies, whereas lower percentages have been achieved with indium 111 or technetium Tc 99m radioconjugates. Even at a usual tumor resolution of 1.5 to 2.0 cm, occult cancers have been disclosed by radioimmunodetection when missed by traditional detection measures, including computed tomography and magnetic resonance imaging. Radioimmunotherapy is somewhat less developed as a treatment modality, but encouraging remissions have been observed, thus stimulating further active pursuit of this technology. These targeting results have been achieved with antibodies that are not truly cancer specific, but only exploit quantitative differences in antigen expression between tumor and adjacent normal tissues. Circulating target tumor antigens do not appear to prevent successful tumor targeting of the radioactive antibodies.