Virtual reality (VR) games has the potential to improve patient outcomes in stroke rehabilitation. However, there is limited information on VR games as an adjunct to standard physiotherapy in improving upper limb function. This study involved 36 participants in both experimental (n = 18) and control (n = 18) groups with a mean age (SD) of 57 (8.20) and 63 (10.54) years, respectively. Outcome measures were the Fugl-Meyer assessment for upper extremities (FMA-UE), Wolf motor function test (WMFT), intrinsic motivation inventory (IMI), Lawton of instrumental activities of daily living (IADL), and stroke impact scale (SIS) assessed at pre-post intervention. The experimental group had 0.5 h of upper limb (UL) VR games with 1.5 h of standard physiotherapy, and the control group received 2 h of standard physiotherapy. The intervention for both groups was performed once a week for eight consecutive weeks. The results showed a significant time-group interaction effect for IMI (p = 0.001), Lawton IADL (p = 0.01) and SIS domain of communication (p = 0.03). A significant time effect was found in FMA-UE (p = 0.001), WMFT (p = 0.001), Lawton IADL (p = 0.01), and SIS domains; strength, ADL and stroke recovery (p < 0.05). These results indicated an improvement in UL motor ability, sensory function, instrumental ADL, and quality of life in both groups after eight weeks of intervention. However, no significant (p > 0.05) group effect on all the outcome measures was demonstrated. Thus, replacing a portion of standard physiotherapy time with VR games was equally effective in improving UL function and general health compared to receiving only standard physiotherapy among stroke survivors.