The major goal of this work is to use human RNases and homologous RNases from other species instead of toxic plant and bacterial proteins in the construction of immunotoxins. Two of the major problems with the clinical use of immunotoxins is the toxicity and immunogenicity of the toxins. The use of human and humanized RNases addresses these problems. Isologous fusion proteins administered to mice are indeed less immunogenic and less toxic than fusion proteins made with plant or bacterial toxins. A human RNase based immunotoxin can prevent the growth of human glioma cells in athymic mice. Furthermore, both single-chain and intact-antibody RNase fusion proteins can be made in the milk of transgenic animals in amounts approaching 1 mg/ml. A homologous RNase (Onconase) from frog oocytes has inherent anti-tumor properties and is in clinical trials as an anti-cancer agent. Onconase has been administered to humans on a weekly basis for up to six months without causing immunological problems or serious toxicities, presumably because of its homology to human plasma RNases. The specificity of Onconase as an anticancer agent can be improved by targeting and to this end the gene has been cloned from Rana pipiens genomic DNA. Furthermore, Onconase has been humanized by making a hybrid protein comprised of sequences from human eosinophil RNase and frog Onconase. Preparation of this protein is in progress.