Although targeted cancer therapies can be enormously effective, most patients eventually develop resistance to them. Even if only a tiny percentage of the original tumor cells contain mutations that make them resistant to a drug, these cells will survive during treatment with this drug, multiply, and eventually produce new tumors. Researchers have created a mathematical model describing this evolutionary process that closely matches clinical findings in patients. The model predicts that giving two or more targeted therapies at the same time will have a better chance of eradicating resistant cells and preventing relapses, because it is unlikely that any given cell will carry multiple mutations that make it resistant to all the drugs being used. However, the model’s predictions still need to be confirmed with actual clinical data.