The action of D-serine as a transmitter molecule in the nervous system has attracted considerable attention in recent years. As a co-agonist required for the activation of N-methyl-D-aspartate receptors (NMDARs), the molecule is thought to play a key role in the mechanism of synaptic plasticity1. However, our ability to study D-serine dynamically in brain tissue has been limited by the lack of an optical biosensor that is non-invasive and specific, such as the Förster Resonance Energy Transfer (FRET)-based biosensor for glycine2. The central issue that has prevented the construction of a similar FRET-based sensor for D-serine is the absence of any naturally occurring, and specific, D-serine binding proteins. To overcome this, we have employed computational protein design to engineer an existing D-alanine/glycine binding protein (DalS)3 towards increased D-serine specificity. Fusion of this engineered binding domain to a pair of donor/acceptor fluorescent proteins created the first genetically encodable FRET-based biosensor for D-serine. Several iterations of computational design and experimental characterisation produced a sensor with high thermostability (Tm = 79 °C) and affinity for D-serine (Kd = 7 μM). This sensor was successfully applied in rat brain slices to visualise changes in hippocampal D-serine levels using two-photon excitation (2PE) fluorescence microscopy. It is hoped that it will become a widely used experimental tool that could yield new insight into the dynamic role of D-serine in synaptic plasticity, learning and memory formation.