In evolutionary genetics, one central aim is to be able to predict aspects of trait evolution. At the very least this requires  an understanding of both how natural selection is acting on a trait, and the genetic architecture of that trait. This gets considerably more complicated when the trait in question is highly multidimensional. In this talk I will discuss our ongoing work to understand why the Drosophila wing is “wing shaped” from both a genetic and evolutionary perspective. Using a 58 dimensional representation of wing shape, I will summarize some of our work identifying the genetic variants that contribute to variation for this trait in natural populations. I will then discuss how natural selection can alter wing shape due to predation, and the potential role of correlated behavioural traits in response to the predators.