How does information processing in neural circuits generate behaviour? Answering this question requires identifying each of the distinct neuronal types that contributes to a behaviour, defining their anatomy and connectivity, and establishing causal relationships between their activity, the activity of other neurons in the circuit, and the behaviour. Here, I propose such an analysis of the neural circuits that guide Drosophila mating behaviours. The distinct mating behaviours of males and females are genetically pre-programmed, yet can also be modified by experience. The set of ~2000 neurons that express the fru gene have been intimately linked to both male and female mating behaviours. This set of neurons includes specific sensory, central, and motor neurons, at least some of which are directly connected. Male-specific fruM isoforms configure this circuit developmentally for male rather than female behaviour. In females, mating triggers a biochemical cascade that reconfigures the circuit for post-mating rather than virgin female behaviour. We estimate that there are ~100 distinct classes of fru neuron. Using genetic and optical tools, we aim to identify each distinct class of fru neuron and to define its anatomy and connectivity. By silencing or activating specific neurons, or changing their genetic sex, we will assess their contributions to male and female behaviours, and how these perturbations impinge on activity patterns in other fru neurons. We also aim to define how a specific experience can modify the physiological properties of these circuits, and how these changes in turn modulate mating behaviour. These studies will define the operating principles of these neural circuits, contributing to a molecules-to-systems explanation of Drosophila s mating behaviours.