Objective: To determine whether there is any difference between virtual reality game-based balance training and real-world task-specific balance training in improving sitting balance and functional performance in individuals with paraplegia. Methods: The study was a pre test-post test experimental design. There were 30 participants (28 males, 2 females) with traumatic spinal cord injury randomly assigned to 2 groups (group A and B). The levels of spinal injury of the participants were between T6 and T12. The virtual reality game-based balance training and real-world task-specific balance training were used as interventions in groups A and B, respectively. The total duration of the intervention was 4 weeks, with a frequency of 5 times a week; each training session lasted 45 minutes. The outcome measures were modified Functional Reach Test (mFRT), t-shirt test, and the self-care component of the Spinal Cord Independence Measure-III (SCIM-III). Results: There was a significant difference for time (p = .001) and Time × Group effect (p = .001) in mFRT scores, group effect (p = .05) in t-shirt test scores, and time effect (p = .001) in the self-care component of SCIM-III. Conclusions: Virtual reality game-based training is better in improving balance and functional performance in individuals with paraplegia than real-world task-specific balance training.