One of the most persistent sources of consternation in the pathway for immuno-oncology drug development today is the high failure rate of early phase trials. Promising treatments, with excellent rationale, and similarly excellent outcomes in model systems are often discarded after failing to demonstrate similar benefit in patients. The difficult translation from preclinical to clinical testing may be in part attributable to heterogeneity in human immunobiology and carcinogenesis which is not recapitulated by animal models. Nonetheless, the contemporary clinical drug development paradigm was designed for traditional cytotoxic drugs and the complexity of modern anticancer therapeutic approaches warrants a rethinking of the most efficient path to transition novel strategies from the laboratory to the clinic. Examples of the shortcomings of current clinical drug development strategies in the context of immunomodulatory therapies include dose escalation designs with drugs that exhibit non-dose dependent effects and often delayed adverse events[1], [2]. Furthermore, there is an enormous unmet need to better understand the mechanisms of action of immuno-oncology drugs, and the underlying mechanisms of resistance that can explain heterogeneity of response and identify biomarkers to help enrich clinical trials for those patients most likely to benefit. To address this need, the platforms on which trials are designed will have to evolve, more seamlessly integrating translational science into studies with clinical endpoints, and changing a traditional ‘bench to bedside’ approach to one of contemporaneous ‘bench and bedside’ interrogation.