Purpose: Given the incidence of stroke, the need has arisen to consider more self-managed rehabilitation approaches. A promising technology is Virtual Reality (VR). Thus far, however, it is not clear what the benefits of VR systems are when compared to conventional methods. Here we investigated the clinical impact of one such system, the Rehabilitation Gaming System (RGS), on the recovery time course of acute stroke. RGS combines concepts of action execution and observation with an automatic individualization of training. Methods. Acute stroke patients (n = 8) used the RGS during 12 weeks in addition to conventional therapy. A control group (n = 8) performed a time matched alternative treatment, which consisted of intense occupational therapy or non-specific interactive games. Results. At the end of the treatment, between-group comparisons showed that the RGS group displayed significantly improved performance in paretic arm speed that was matched by better performance in the arm subpart of the Fugl-Meyer Assessment Test and the Chedoke Arm and Hand Activity Inventory. In addition, the RGS group presented a significantly faster improvement over time for all the clinical scales during the treatment period. Conclusions. Our results suggest that rehabilitation with the RGS facilitates the functional recovery of the upper extremities and that this system is therefore a promising tool for stroke neurorehabilitation.