Philosophers of science have overlooked the role of theory in neuroscience, resulting in a somewhat surprising naïveté regarding the nature and function of neuroscientific theories. Here I provide a framework that identifies and begins to characterize what we need to know about neuroscientific theories so as to improve our epistemic standing. I argue that we need an account of the structural, interpretive, and functional aspects of neuroscientific theories, using the neuron doctrine as an illustrative case study. I introduce the novel metaphor of theoretical infrastructure as a guide for making sense of neuroscientific theories and their place within neuroscientific practice.