One of the most enduring problems of evolutionary theory and biology in general is how can complex organisms arise from random genetic change. This is also the ultimate question guiding the proposed research on developmental evolution in response to natural selection. Because of the complexity of the developmental process, most mutations have a variety of effects, but only a few of them are likely to be adaptive. For that reason any adaptive evolutionary change is a complex mix of adaptive, fitness increasing effects, deleterious side effects and compensatory changes ameliorating these deleterious side effects. This is the likely reason why evolution at the genomic and developmental level is exceedingly complex. Now, for the first time in history, with systems biology combined with novel methods from quantitative genetics we have the tools to probe and eventually understand the complexities of evolutionary change at the genetic and developmental level. In this project, I propose to explore the idea of developmental evolution by evolving pleiotropic genes, accompanied with the compensation of side effects. I will apply this model on the specific question, namely the differentiation between fore and hind limb in mice. Specifically, the idea is that homologous parts share the crucial developmental program while its expression becomes modified by the local factors. The developmental program and its modifying background maintain functionality by coadaptation, in this case allowing the two limb pairs to diverge in function. Using existing population variation in this mechanism, I will use a novel quantitative genetic approach to map the factors underlying differentiated development of the two limb pairs. The data used for this project stem from mouse intercross of two inbred lines. I attempt to complement computational approach with experimental validation of expression patterns. The model provides a unifying mechanism for evolutionary differentiation of parts at different levels. Within the population this model may underlie differentiation of traits in sexual dimorphism, above the population level it may underlie speciation by divergence of coadapted interacting gene complexes. This project provides a direct link between developmental and population genetics. Combined computational and empirical results will contribute towards the more realistic theoretical models.