This chapter is concerned with the large-scale structure of the human brain, how that relates to human behaviour, and how certain types of modularity could have emerged through evolution as a result of a combination of fitness advantages and neurophysiological constraints. It is argued that computational modelling involving evolutionary neural network simulations is the best way to explore such issues, and a series of simulation results are presented and discussed. The incorporation of known neurophysiological constraints into the models is seen to be of crucial importance.