Recent advances in cytotoxic therapies for pancreatic ductal adenocarcinoma (PDA) are overshadowed by stalled clinical progress of more targeted strategies, the vast majority of which have failed in clinical trials. Inability to translate preclinical promise into clinical efficacy derives, in part, from imperfect disease modeling and mismatches between preclinical and clinical study design and execution. Into these gaps fall our patients who enter the clinical trial landscape expectantly and bear the brunt of its inadequacies. If improving patient survival is paramount, then it must be acknowledged that the failure of a phase III trial represents a larger failure of all of the work that preceded it. Repeated failures suggest a need to reappraise the current preclinical-to-clinical apparatus. Exceptional models of PDA are now available to researchers, and the first steps toward a new era of success can begin with improved selection and application of these systems. We discuss the key features of the major preclinical platforms for PDA and propose a paradigm for rigorous interrogation of prospective therapies.