The successful introduction of genetically engineered human and chimeric immunoglobulin proteins has established monoclonal antibodies (mAbs) as a validated approach for treating malignancies. The unique properties of mAb therapies including their high affinity and specificity, and the differential expression of target antigen in tumor cells versus normal cells make them attractive agents for cancer immunotherapy. The field of immunoconjugate development attempts to combine the specificity of mAb therapies with cytotoxic and radionuclide molecules, thereby combining the best characteristics of these two different modalities. Two radiolabeled mAbs, (90)Y-ibritumomab tiuxetan and (131)I-tositumomab, and one drug conjugate, gemtuzumab ozogamicin have been approved for the treatment of malignancies. Other conjugates carrying toxic payloads of calicheamicin, geldanamycin, maytansinoids and taxoids as well as peptide exotoxins are undergoing preclinical and clinical development. Nevertheless, several obstacles have limited robust antitumor activity and broad application of imunoconjugates, including the optimization of three structural components of the immunoconjugate (i.e. mAb and target specificity, chemical linker design, and the cytotoxin), as well as issues common to mAb therapy such as heterogeneous antigen expression, which can limit uniform antibody delivery. This Review examines optimal design, the lessons learned from clinical immunoconjugate development, and the promising agents in early preclinical/clinical development for the treatment of cancer.