Social facilitation and social support literature, diverging with regards to increasing versus decreasing of an individual’s tension, apprehend different aspects of “the presence of others.” To examine the neural correlates of social presence effects, whether “the presence of others” increases or decreases an individual’s tension, we measured prefrontal activation while participants performed a driving video game task using near-infrared spectroscopy (NIRS). Participants were divided into single and paired groups, and then sub-divided according to their game proficiency (high and low). The participant’s task was to drive from start to goal with a default route map without an observer (single group) or under observation by an acquainted partner (paired group). The paired participants alternated their player-observer roles in a turn-taking style (Driver first and Observer second: D1-O2; Observer first and Driver second: O1-D2). The behavioral data demonstrated that, regardless of game proficiency, D1 in the paired group yielded fewer errors and longer driving time than single players, while no differences were found between D1 and D2. The tension evaluation scores in single players and D2 were higher than D1. In turn, the NIRS data revealed that, in low-proficiency players, single players and D2 who first observed D1’s performance showed higher activation than D1, but neither did so in high-proficiency players. These results suggest that the presence of an acquainted partner (O1) functions positively to reduce an individual’s (D1) tension in low-proficiency players. However, prior observation of another’s performance may negate the positive social presence effect leading to an increase of tension in the subsequent task.