Gross motor skills (GMS) are the foundation for humans reaching an optimum level of motor competence necessary to undergo normal development, maintain health, and achieve athletic excellence. Yet, there is evidence that GMS levels of children are on a decline globally. Therefore, the main purpose of this study was to investigate the effectiveness of traditional cultural games (TCG) skills, practiced according to different amounts of contextual interference (CI), on the acquisition and retention of GMS. A total of 103 Pakistani primary school children aged between 7 and 10 years were recruited for this study. Participants were randomly assigned to four practice groups with different amounts of CI: Block (B) (low interference), gradually increasing (GI) (moderate interference), random (R) (high interference), and game-based (high interference). The Test of Gross Motor Development (TGMD-2) was used to assess four tasks [overhead throw (OT), underhand throw (UT), catch (C), and throwing to a target]. The test was administered on four occasions: during pre-test, post-test, retention, and transfer. The results showed that the R group outperformed all the other groups in the post-test and the retention test. Meanwhile, in the transfer test, both R and Game-Based groups performed better than the B and GI groups. There were no differences between the R and Game-Based groups during transfer. Practicing TCG skills according to a random order was better for the acquisition and learning of GMS. The CI effect was evident, whereby high interference practice schedules were superior to low and moderate interference practice schedules.