There is a lack of studies that examine the longitudinal relationship between physical activity (PA) and mental health with accelerometers in adolescent populations. The current study aimed to investigate whether changes in PA intensities and sedentary time (ST) are associated with mental health in a cohort of adolescents. 246 (130 boys, 116 girls) high school adolescents aged 13-15 years old (13.28 ± .57) were recruited in a city from the north of Spain. PA and ST levels were assessed by accelerometry. Psychological well-being (PWB) was measured using a self-reported validated questionnaire. Regression linear models were implemented by including PWB as the dependent variable and PA intensities and ST as predictors. Girls who increased light, moderate, and vigorous PA had significantly better PSW at follow-up, whereas no longitudinal associations were found for boys. No association between ST and PSW was identified. The results suggest changes in PA, but no in ST, have an association with PWB in adolescents. Nevertheless, the associations were stronger in girls than in boys. Because girls showed worse PWB at both time points, an optimal threshold for PA influence on mental health could explain the results. Future research is warranted.