Transcriptional assays such as yeast two hybrid, split ubiquitin, and Tango that convert transient protein-protein interactions (PPIs) in cells into stable expression of transgenes are powerful tools for PPI discovery, high-throughput screens, and analysis of large cell populations. However, these assays frequently suffer from high background and they lose all information about PPI dynamics. To address these limitations, we developed a light-gated transcriptional assay for PPI detection called PPI-FLARE (PPI-Fast Light-and Activity-Regulated Expression). PPI-FLARE requires both a PPI to deliver TEV protease proximal to its cleavage peptide, and externally-applied blue light to uncage the cleavage peptide, in order to release a membrane-tethered transcription factor (TF) for translocation to the nucleus. We used PPI-FLARE to detect the ligand-induced association of 12 different PPIs in living mammalian cells, with a temporal resolution of 5 minutes and a ±ligand signal ratio up to 37. By systematically shifting the light irradiation window, we could reconstruct PPI time-courses, distinguishing between GPCRs that engage in transient versus sustained interactions with the cytosolic effector arrestin. When combined with FACS, PPI-FLARE enabled >100-fold enrichment of cells experiencing a specific GPCR-arrestin PPI during a short 10-minute light window over cells missing that PPI during the same time window. Due to its high specificity, sensitivity, and generality, PPI-FLARE should be a broadly useful tool for PPI analysis and discovery.Protein-protein interactions (PPIs) are central to cellular signal transduction. Consequently, many assays have been developed to detect and study them, particularly in the context of living cells, where native PPIs are unperturbed by cell lysis, detergents, fixatives, or dilution. For example, FRET 1 , BRET 2 , fluorescence correlation spectroscopy (FCS) 3 , protein complementation assays (PCAs) 4 , and fluorescence relocalization assays 5,6 have all been used to visualize PPI dynamics in living cells. Though information-rich, these assays have the downside of being labor-intensive, requiring high-content or time-lapse microscopy, and are consequently difficult to perform on a large scale. This makes them non-optimal for PPI discovery, for adaptation to high-throughput screens, or for analysis of large cell populations such as in complex tissue. Instead, real-time assays are more suited for the focused study of a small number of known PPIs under a small set of conditions or in a small number of cells.A separate class of assays detects PPIs by signal integration rather than real-time imaging, and produces gene transcription as the readout. Examples include the yeast two hybrid assay 7 , the split ubiquitin assay 8 , and Tango 9-11 . Benefits of these assays include scalability (because real-time microscopy is not needed), versatility of read out (transcription of a fluorescent protein or an antibiotic resistance gene, for example), and high sensitivity due to signal amplification. Thes...