Background: Visual and somatosensory integration processing is needed to reduce pusher behavior (PB) and improve postural control in hemiplegic patients with acute stroke. Objects: This study aimed to investigate the effects of game-based postural vertical training (GPVT) on PB, postural control, and activity daily living (ADL) in acute stroke patients. Methods: Fourteen participants with acute stroke (<2 months post-stroke) who had PB according to the Burke lateropulsion scale (BLS) (score>2) were randomly divided into the GPVT group (n1=7) and conventional postural vertical training (CPVT) group (n2=7). The GPVT group performed game-based postural vertical training using a whole-body tilt apparatus. while the CPVT group performed conventional postural vertical training to reduce PB (30 minutes/session, 2 times/day, 5 days/week for 3 consecutive weeks). The BLS was evaluated to assess the severity of PB. And each subject’s postural control ability and ADL level were assessed using the postural assessment scale for stroke (PASS), balance posture ratio (BPR), and Korean-modified Barthel index (K-MBI). Outcomes were measured preand post-intervention. Results: Comparison of the pre- and post-intervention assessment results showed that both interventions led to the following significant changes: decreased severity of PB scores and increased PASS, BPR, and K-MBI scores (p<.05). In particular, statistical analysis between the two groups, the BLS score was significantly decreased in the GPVT group (p<.05). And PASS, BPR, and K-MBI scores were significantly improved in the GPVT group than in the CPVT group (p<.01, respectively). Conclusion: This study demonstrated that GPVT lessened PB severity and improved postural control ability and ADL levels in acute stroke patients.