PROJECT SUMMARY Understanding how information is processed in the mammalian neocortex has been a longstanding question in neuroscience. While the action potential is the fundamental bit of information, how these spikes encode representations and drive behavior remains unclear. In order to adequately address this problem, it has become apparent that experiments are needed in which activity from large numbers of neurons can be measured in a detailed and comprehensive manner across multiple timescales. Direct measurements of action potentials have primarily been achieved by electrophysiology. However, such measurements cannot easily be combined with other methods to assess the connectivity and molecular properties of neurons. Integrating functional, anatomical, and genetic information is critical for understanding how neuronal circuits are organized and computed. There have been long-standing efforts in developing optical methods for measuring neuronal activity due to its compatibility to simultaneously measure connectivity and molecular identity using fluorescent labeling techniques. We have developed a two-photon-excitable genetically-encoded voltage-sensitive indicator and ultra-fast two-photon microscope that enables optical measurements of action potentials deep into the brain. However, imaging at high signal-to-noise beyond several minutes remains challenging due to photo-bleaching and risks of photo- damage. In order for these new technologies to become more robust for neuroscience applications, it is necessary to improve upon the stability, reliability, and efficiency of two-photon voltage imaging. To achieve this, it requires a concerted effort between optical engineers, protein engineers, and computational scientists to optimize instrumentation, sensors, and image analysis for broad dissemination. This multi-investigator effort proposes to advance two-photon voltage imaging to enable sustained tracking of population activity at timescales of animal behavior and learning.