Waddington is usually acknowledged as a biologist who proposed a more subtle concept of environment than the one generally in currency during the rise of the Modern Synthesis. As such, he was among the few scientists in the mid-twentieth century to develop an elaborated concept of the environment that would fully embrace its constructive role both in development and evolution. Yet, on close inspection, there is an inconsistency in Waddington's theoretical positioning. On the one hand, as a critic of population genetics, Waddington never stopped claiming that natural selection acts on phenotypes and that the phenotype is the outcome of both the genome and the environment. On the other, however, the topology of his famous epigenetic landscape was anchored only in the genome, and the variation of the environment was treated as an external perturbation. In other words, the genes and the environment were sometimes considered as symmetric agents in the epigenetic system, and sometimes not. My aim is to shed light on the significance of this tension in Waddington's theoretical framework. I show that Waddington's biology is best characterized as an asymmetric understanding of the causal role of genes and environment both in development and evolution. His model of genetic assimilation was based on the idea of differential levels of hereditary responsiveness to environmental variations, giving the genome a leading role and paving the way for critics of his conceptions. In the final section, I argue that eventually Waddington was trapped by his diagram of the epigenetic landscape, which might also have been an obstacle to achieving a proper conception of the creativity of embryogenesis.