Immunotherapy, based on mAbs specifically directed against cancer cells, is considered a precious strategy in the fight against cancer because of its selectivity and lack of multidrug resistant effects. However, there are obstacles to the complete success of current immunotherapy such as immune responses to nonhuman or even humanized antibodies and the large size of the antibodies, which hinders their diffusion into bulky tumors. Fully human, small immunoagents, capable of inhibiting tumor growth may overcome these problems and provide safe, highly selective and effective antitumor drugs. An attractive target for immunotherapy is ErbB2, a transmembrane tyrosine kinase receptor, overexpressed on tumor cells of different origin, with a key role in the development of malignancy. An anti-ErbB2 humanized monoclonal (Herceptin) is currently used with success for breast cancer therapy; however, it can engender cardiotoxicity and a high proportion of breast cancer patients are resistant to Herceptin treatment. Anti-ErbB2 immunoagents of human origin, with potentially no or very low immunogenicity have been engineered to assemble 'compact', i.e. reduced size, antibodies, one consisting of a human single-chain antibody fragment (scFv) fused to a human RNase to construct an immunoRNase and the other made up of two human scFv molecules fused to the Fc region of a human IgG1. By choosing a human antibody fragment as the immune moiety and a human RNase as the effector moiety, an immunoRNase would be both nonimmunogenic and nontoxic, as it becomes toxic only when the scFv promotes its internalization by target cells. The alternative strategy of compact antibodies was aimed at producing therapeutic agents with an increased half-life, prolonged tumor retention and the ability to recruit host effector functions. Moreover, the bivalency of compact antibodies can be exploited to construct bispecific antibodies, as well as for other therapeutic applications.