Synthetic biology is advancing rapidly as biologists, physicists and engineers are combining their efforts to understand and program cell function. By characterizing isolated genetic components or modules, experimentalists have paved the way for more quantitative analyses of genetic networks. A recent paper presents a method of computational, or in silico, evolution in which a set of components can evolve into networks that display desired behaviors. An integrated approach that includes a strategy of in silico design by evolution, together with efforts exploiting directed evolution in vivo, is likely to be the next step in the evolution of synthetic biology.