Upper-limb virtual rehabilitation (VR) in adult acquired brain injury (ABI) is based largely on systems administered on a one-to-one basis. Multi-user interaction between co-located participants may offer advantages over single-user methods. The present study examined the feasibility of deploying a co-located VR system (Resonance) in a clinical setting. Following a baselining period, 5 patients with ABI completed 12 Resonance sessions over 4–6 weeks. Feasibility criteria included recruitment, intervention delivery, attrition, user experience, and suitability of outcome measures. Individual participant motor proficiency (box and blocks task) was examined using a time-series analysis with reliable change indices and curve fitting. All feasibility criteria were satisfied, with positive reports of user experience. Repeated collection of outcome measures was successfully integrated into the training schedule. Time-series analysis was successfully conducted, providing a detailed account of individual training-related change. Within a clinical setting, it was feasible to deliver Resonance and regularly monitor motor function. User feedback regarding the co-located VR intervention was generally positive, but expectations regarding the level of immersion may need to be managed. Individual time-series analysis is recommended as an adjunct to group-based analysis in future VR research. These findings can inform the design of a clinical trial.