Background and purpose: In recent years, sedentary behaviour (SB) has become a significant risk factor for health, alongside the decline of physical activity (PA) in the population. This study explored adolescents’ PA and SB using a longitudinal design, from the age of 13 to 15, to elucidate any changes and when they potentially occur. Method: The data material from the accelerometer measurements was analysed using the program SPSS, version 28. Repeated measures ANOVA (and follow-up test with Bonferroni corrections) were used to assess changes in the four activity levels during the three years. Independent t-tests were used to examine gender differences in each year and at each activity level and to determine differences in activity level between participants with valid data for all three years and those with valid data only at 13 and/or 14 years of age. Results: Statistical analyses of the accelerometer measurements revealed a significant increase in time in the lowest activity level and a decrease in light, moderate and vigorous intensity PA levels during the three years. These changes were larger from 13 to 14, than from 14 to 15, years of age. However, no significant interaction was found between time and gender. Analyses also showed a significantly higher time spent in SB and lower time spent in moderate activity among girls compared to boys at 13, but not at the age of 14 and 15. Conclusions: The findings, which indicate a reduction in PA and increased sedentary time with increasing age, are discussed, which we argue are important to increase and solidify involvement in PA among adolescents. Emphasized here is the importance of communicating the diverse possibilities of movement and challenging understandings of what PA can be for youth.