The observation that different species within one animal phylum resemble one another in the course of embryonic development fascinated scientists since the dawn of developmental biology. Already in 1828, Carl Ernst von Baer formulated his third law of embryonic development, later extended into the hourglass model that recognizes the phylotypic stage during mid-embryogenesis as the most constrained period of development, preceded and followed by greater morphological divergence. The arguments for the existence of the developmental hourglass come primarily from comparative morphological analysis of vertebrate embryos and are, to the large extent, qualitative.This proposal describes experiments designed to test the hourglass model quantitatively using sequenced Drosophila species as a model system. The approach is unique, because by focusing on a set of highly similar species, it will be possible to precisely quantify the differences among them and use statistical and genetic modeling approaches to evaluate the hourglass hypothesis.The divergence of genome expression and embryo morphology will be evaluated before, during and after the phylotypic stage, using deep sequencing and high-resolution microscopy combined with advanced image analysis. The two sets of measurements will be linked through cross-species, reverse genetic manipulation of the genomes, attempting to establish a causal chain of events from sequence divergence, through divergence in gene expression patterns to the divergent phenotype of the embryo.The quantitative analysis of the developmental hourglass will provide conclusive evidence for or against the existence of the constraints in gene regulatory networks underlying the phylotypic stage and shed light on one of the most fundamental laws in biology.