A perennial debate in the philosophy of psychology is the debate over the autonomy of the discipline from neuroscience. Traditionally the debate has taken place over a metaphysical question: are mental states multiply-realized by brain states? I take it that this tradition is misleading. The multi-realizability question, though genuine and interesting in its own right, has no bearing on the question of the autonomy of psychology from neuroscience. The aim of this talk is to provide a defense of a genuine non-autonomy thesis, a normative epistemological one: when choosing between different psychological theories, psychologists should make use of neuroscientific findings.