Evolution is the proximate cause of death of many and perhaps most cancer patients. Many effective drugs for metastatic and/or advanced-stage cancers have been developed over the past decade, although the evolution of resistance remains the major barrier to disease control or cure. In large, diverse populations such as the cells that compose metastatic cancers, the emergence of cells that are a priori resistant or can rapidly develop resistance is virtually inevitable. Attempts to disrupt the molecular mechanisms of resistance have generally been unsuccessful in clinical practice. An alternative approach focuses on controlling the Darwinian processes driving the eco-evolutionary dynamics of treatment-resistant cancer populations. That is, we hypothesize the genetic and epigenetic mechanisms governing emergence of treatment resistance within a cancer population cannot be prevented. However, significant resistance occurs only when the initially small population of resistant cells proliferates to form a tumor of sufficient size to impact a patient’s clinical course. This process is governed by eco-evolutionary dynamics. Mathematical models based on evolutionary first principles allow clinical oncologists to anticipate and steer the evolutionary dynamics of treatment-sensitive and treatment-resistant cancer cells. Multiple mathematically guided evolution-informed clinical trials have been completed and many are ongoing. Equally important, when model parameters are updated using longitudinal trial data, computer simulations can be used to visualize the underlying evolutionary dynamics that led to the observed outcome in each trial patient and identify alternative strategies that would have optimally prolonged survival.