Model-based estimates of tumor growth inhibition (TGI) metrics have the potential to enhance learning in early (phase II) clinical studies. They can be used as end points and biomarkers to predict treatment effect on clinical outcome measures-e.g., overall survival (OS)-and support phase II study design, end-of-phase II decisions, and phase III planning and execution. Efforts should be made to assess models in simulating independent studies with treatments of varying mechanisms of action.