In spite of recent progress, our understanding of cognitive disorders remains tenuous. While outward symptoms of neurodegenerative and mental disorders, ranging from Alzheimer’s to schizophrenia, are readily apparent, their underlying cellular mechanisms are unclear. Most, if not all, exhibit some form of synaptic dysfunction and/or circuit abnormality. Unfortunately, our ability to monitor disruptions in synapse or circuit connectivity as they occur in vivo has been hindered by the difficulty of visualizing individual synaptic contacts at sufficient resolution to discern their formation or elimination. The primary challenge for imaging synaptic connections lies in the narrow cleft separation of 20-50 nm that is far below optical resolution. Here we propose a new approach to identify synapse formation and dissociation in vivo by monitoring the distance between pre- and post-synaptic protein pairs using fluorescence resonance energy transfer (FRET). In Specific aim 1 we will develop wide-field depth-resolved FLIM-FRET by implementing De-scattering with Excitation Patterning (DEEP), a wide-field depth resolved imaging approach based on structured light temporal focusing two-photon excitation that we recently demonstrated is compatible with in vivo neural imaging. We propose implementing DEEP for FLIM by using avalanche photodiode arrays with nanosecond gating to simultaneously resolve lifetimes with over two thousand detectors. For testing microscope development, we will express in the mouse brain, in vivo, known intramolecular FRET pairs using our previously developed methods for sparse, multi-fluorophore neuronal labeling. In Specific aim 2 we will generate a series of FRET donor/acceptor molecules fused to variants of the neuroligin-neurexin trans-synaptic partners in a variety of configurations, designed so that donor-acceptor distance is kept within ~5 nm in the bound state for FRET to occur. These fusion constructs will be screened in cultured neurons and selected based on faithful synaptic localization, lack of interference to normal synaptic dynamics, and the presence of strong FRET signal upon fusion partner binding. The in vivo labeling strategy will be a modification of one we recently developed for imaging Layer 2/3 pyramidal cell dendritic arbors and their resident synapses in vivo using a three-color two-photon system, modified to avoid co-expression of donor and acceptor in the same cell. The postsynaptic fusion protein will be co-expressed with a cell fill to visualize a single targeted cell with all its postsynaptic sites. Where these sites contact labeled presynaptic terminals, transsynaptic binding should place the fluorescent donor/acceptor pairs in close proximity, allowing FRET. Selected pairs will then be tested in vivo in the brain for performance in the presence of autofluorescence and signal loss from scattering in deep layers.