To turn sensory inputs of the outside world into appropriate actions, our brains as well as the brain of all animals must perform specific computations. These computations can be simple or complex, but often involve intricate circuit architecture. It is our major goal to understand how neural computations are implemented in the neuronal networks for the brain. We chose to study a basic, paradigmatic computation, the extraction of direction-selective signals, which is a hallmark of motion detection. While recent years have mapped out core circuits of visual motion detection in the model organism Drosophila, and highlighted many parallels to vertebrate vision, we now aim to understand circuit computations at a new level of detail. Within the “MicroCyFly” project, we develop novel genetic tools to dissect visual circuitry at the level of individual neuronal connections, or synapses, rather at the cell type level. To understand how microcircuits in the brain are generally organized to perform certain tasks is key to target many neurological and psychiatric diseases.