Biometric analyses of variability in behavioral phenotypes have demonstrated that genotype plays a significant role in all behavioral development, but the developmental significance of the environment has remained obscure. Behavior genetic analyses typically show the effect of shared family environment to be very small, although considerable variability remains to be explained after genetic factors have been accounted for. Behavior geneticists have suggested, contrary to intuition, that almost all of the important effects of the environment serve to make family members more different from each other. Environmentalists of several persuasions have pointed to the crucial importance of the environment in the initiation and regulation of all developmental processes. In this article, we present a series of simulations to suggest that some of the difficulty of identifying environmental effects in biometric models is methodological. Adding simple dynamic parameters to models of development leads to systems in which environment produces substantial variability that can be detected in the context of a particular genotype, but vanishes when genotype is allowed to vary.