The substantia nigra pars reticulata (SNr) is a primary output for basal ganglia signalling. It plays an important role in motor control, integrating inputs from upstream structures in the basal ganglia, before sending organised projections to a wide range of targets in the midbrain, brainstem and thalamus. Here we present a detailed in silico model of the mouse SNr, including its afferent inputs. The electrophysiological and morphological properties of SNr neurons are characterised in acute brain slices via whole cell patch-clamp and morphological reconstruction. Using reconstructed morphologies, multicompartmental single-cell models are instantiated within the NEURON simulation environment and populated with appropriate ion channel models. Model parameters are optimised via an evolutionary algorithm, such that simulated neurons faithfully reproduce recorded electrophysiological behaviour. Using the simulation software Snudda, neuron models are incorporated into a circuit-level model, where the sparse collateral connectivity within the SNr can be recreated. We simulate the mouse SNr at scale, featuring realistic volumes and neuron density. The unique synaptic properties of different afferent sources are captured in silico, with spatially organised connectivity established via touch-detection. Simulated afferent activation reproduces several experimental observations. Furthermore, we employ this model to predict afferent connectivity patterns at the cellular level, as well as SNr responses to competing inputs. The model integrates into a previously-described large-scale model of the dorsal striatum, for further investigation into basal ganglia information processing (Hjorth et al., 2020).Reference: Hjorth et al., 2020, Proc. Natl. Acad. Sci. U. S. A. 117(17):9554-9565.