In principle, cell surface receptors that are overexpressed in tumor tissue could serve as targets for anticancer drugs attached to receptor ligands. The purpose of this paper is to identify the necessary elements for a successful receptor-targeted drug. We used the gastrin/cholecystokinin type B receptor as a model delivery system, and we report on the synthesis, trafficking, and in vitro and in vivo evaluation of heptagastrin, the C-terminal heptapeptide of gastrin, linked via an appropriate linker to a potently cytotoxic ellipticine derivative, 1-[3-[N-(3-aminopropyl)-N-methylamino]propyl]amino-9-methoxy-5, 11-dimethyl-6H-pyrido[4,3-b]carbazole. These data, and previous work from our laboratory, show that the drug-complexed ligand is sorted to lysosomes whereas the receptor is recycled to the plasma membrane. The lysosomal processing of the ligand/drug construct depends on the linker between the ligand sequence and the cytotoxic moiety. We show that heptagastrin linked to ellipticine via a succinoyl-substituted pentapeptide, AlaLeuAlaLeuAla, is at least 10(3) more toxic to cholecystokinin type B receptor-positive NIH/3T3 cells than to isogenic NIH/3T3 cells lacking the receptor. The conjugated drug eradicated all receptor-positive tumor cells in vivo without producing any general toxicity. The data indicate that the density of the cell surface receptor, the properties of the cytotoxic moiety, and the correct processing of the drug-conjugated ligand in lysosomes are crucial to the effectiveness of a receptor-targeted drug.