BACKGROUND: We investigated prospective associations between physical activity/sedentary behaviour (PA/SED) and General Certificate of Secondary Education (GCSE) results in British adolescents. METHODS: Exposures were objective PA/SED and self-reported sedentary behaviours (screen (TV, Internet, Computer Games)/non-screen (homework, reading)) measured in 845 adolescents (14·5y ± 0·5y; 43·6 % male). GCSE results at 16y were obtained from national records. Associations between exposures and academic performance (total exam points) were assessed using multilevel mixed-effects linear regression adjusted for mood, BMI z-score, deprivation, sex, season and school; potential interactions were investigated. RESULTS: PA was not associated with academic performance. One-hour more accelerometer-assessed SED was associated with (β(95 % CI)) 6·9(1·5,12·4) more GCSE points. An extra hour of screen time was associated with 9.3(-14·3,-4·3) fewer points whereas an extra hour of non-screen time (reading/homework) was associated with 23·1(14·6,31·6) more points. Screen time was still associated with poorer scores after adjusting for objective PA/SED and reading/homework. CONCLUSIONS: An extra hour/day of screen time at 14·5y is approximately equivalent to two fewer GCSE grades (e.g., from B to D) at 16y. Strategies to achieve the right balance between screen and non-screen time may be important for improving academic performance. Concerns that encouraging more physical activity may result in decreased academic performance seem unfounded.