Abstract: Evolutionary biologists have long sought to infer historical patterns of relatedness among organisms using phylogenetic methods, and to gauge the evolutionary processes which determine variation among individuals in extant populations. Relatively less effort has been devoted to developing evolutionary biology as a truly predictive science, where we can gain the power to foresee the course of future evolutionary events. Accurate predictions of evolvability would be particularly useful in the evolution of infectious diseases, because we would be better able to meet the challenges created by pathogens evolved to newly emerge in humans and other populations. Experimental evolution of microbes in the laboratory offers the possibility to rigorously test hypotheses relating to pathogen evolvability. Here we review results from recent experiments using RNA viruses. These studies suggest that adaptability of virus populations in new environments can be predicted if knowledge exists of the genetic architecture underlying viral traits in the current environment. Knowledge of genetic robustness (phenotypic canalization in the face of mutational input) may be sufficient for making accurate predictions of virus evolvability; however, experiments reveal that robustness may either promote or hinder evolvability depending on the particulars of new environmental challenges. Additional studies show that viruses historically evolved on multiple hosts tend to acquire molecular substitutions fostering generalized host- use, allowing accurate predictions of the relationship between known host-range and the ability for viruses to emerge on novel hosts. Although the process of evolution is blind, these studies offer hope that the possibility of converting evolutionary biology into a predictive science is not as gloomy as we once feared.