Resistance to targeted therapies is commonplace, so should we use more than one drug and target different pathways? Martin Nowak and colleagues built a mathematical model to address this question. They used lesion regression and growth data from 20 patients with metastatic melanoma who were treated with vemurafenib to generate the model parameters. They then applied the model to independent data sets from patients with metastatic disease. According to the model, dual therapy with two hypothetical drugs that target two different pathways would result in long-term control, as long as there are no single mutations that produce cross-resistance to both drugs. Patients with substantial disease burden would require three or more drugs for long-term benefit. Moreover, the model shows that combined rather than sequential treatment is far better.