Despite the importance of physical education (PE) lessons for physical activity in adolescents, the acute cognitive responses to PE lessons have not been explored; a gap in the literature that this study addresses. Following familiarisation, 76 (39 female) adolescents (12.2 ± 0.4 y) completed two trials (60 min games-based PE lesson and 60 min academic lesson) separated by 7-d in a counterbalanced, crossover design. Attention, executive function, working memory, and perception were assessed 30 min before, immediately post, and 45 min post-lesson in both trials. Participants were split into high-and low-fit groups based on a gender-specific median split of distance run on the multi-stage fitness test. Furthermore, participants were split into high and low MVPA groups based on a gender-specific median split of MVPA time (time spent >64% HR max) during the PE lesson. Overall, a 60 min games-based PE lesson had no effect on perception, working memory, attention, or executive function in adolescents (all p > 0.05) unless MVPA time is high. The physical activity-cognition relationship was moderated by MVPA, as working memory improved post-PE lesson in adolescents who completed more MVPA during their PE lesson (time*trial*MVPA interaction, p < 0.05, partial η2 = 0.119). Furthermore, high-fit adolescents displayed superior cognitive function than their low-fit counterparts, across all domains of cognitive function (main effect of fitness, all p < 0.05, partial η2 0.014–0.121). This study provides novel evidence that MVPA time moderates the cognitive response to a games-based PE lesson; and emphasises that higher levels of fitness are beneficial for cognitive function in adolescents.