The neural circuits that underlie motion vision in Drosophila provide an excellent model system for studying the logic of neural computation. A rich history of quantitative behavioral analysis has provided a detailed theoretical framework for investigating the neuronal basis of motion detection. Here, we describe the many inventive methods that have been used to quantify optomotor responses in Drosophila. Next, we discuss stimuli that can probe the computational structure of motion detection circuitry. Finally, we highlight some of the sophisticated genetic tools for targeting and manipulating neurons that are being applied to the visual system. Studies combining quantitative behavioral assays with this ever-expanding genetic toolkit are beginning to uncover the computational roles played by individual neurons. The circuits that process visual motion in Drosophila provide an exciting opportunity to understand a complete neural circuit linking perception to behavior.