There is a critical unmet need in cancer research to identify targeting agents that can distinguish between normal, pre-cancerous, malignant, and metastatic phenotypes. The purpose of his proposal is to identify functional ligands that can specifically bind and internalize into cancerous 3rostate cells. We will employ our new ligand selection methods specifically designed for complex living targets. These selection strategies eliminate common problems associated with selection on complex cellular targets, and allows for the rapid identification of internalizing ligands from complex synthetic or natural peptide libraries. We will use these technologies to identify functional ligands that internalize into benign and malignant human prostate cells obtained from clinical tissue specimens. Together, these screens will allow us to begin identifying ligands that distinguish normal and malignant prostate tissue. Ligands identified in these studies have many direct commercial applications in the treatment of prostate cancer including the targeting of cytotoxic drugs. The ligands may also be useful for basic research on the biology of cancer