Phenotypic evolution depends on heritable variation in phenotypes. A central aim of evolutionary biology, therefore, is to understand how processes generating phenotypic variation interact with selection and drift to result in phenotypic evolution. Recent studies have highlighted the propensity for populations to harbor genetic variation that contributes to phenotypic variation only after some environmental or genetic change. Many authors have suggested that release of this cryptic genetic variation by stressful or novel environments can facilitate phenotypic adaptation. However, there is little empirical evidence that stressful or novel environments release cryptic genetic variation, or that, once released, it contributes to phenotypic evolution. We argue that empirical studies are needed to answer these questions, and identify the empirical approaches needed to study the relationship between environment, released cryptic genetic variation and phenotypic evolution.