The level of physical activity in youngsters worldwide is declining (Kohl et al., 2012). Peers are an important determinant of physical activity in youngsters (McPherson, Smith-Lovin, & Cook, 2001) This study is particularly interested in the processes that explain the similarity in physical activity in adolescents’ friendships. More specifically, this study will examine changes in friendships and physical activity levels at three time points and test whether similarity in physical activity exists prior to the formation of friendships (selection), or whether friends become more similar over time (influence). To test for selection and influence effects of friendship networks, stochastic actor-based models will be used (Snijders, van de Bunt, & Steglich, 2010). Longitudinal social network analyses simultaneously investigate changes in friendships and physical activity, and are capable of disentangling whether friends select other’s based on physical activity and whether friends influence each other’s physical activity. In total, 394 youngsters (41% boys, age: 11-14 y/o) out of eight secondary school classes participated for three separate weeks between January and June of 2016. Physical activity was measured by wrist-worn accelerometer (Fitbit Flex) for five days (M=8424.06 steps/day, SD =3595.37). Friendship was measured by asking participants to nominate their friends, via the research smartphone. Social network modelling revealed that, after controlling for network and same-sex selection effects, the selection effect was statistically significant (b=2.71, SE= 1.13, p=.017). This indicates that youngsters initiate friendships with others that have similar physical activity levels. In addition, a statistically significant influence effect was observed (b=.34, SE= .14, p=.017), which indicates that friends became increasingly similar in physical activity. The findings provided evidence for similarity in physical activity in youngsters. Youngsters who were not friends at the beginning of the study, but were similar in their physical activity level, were more likely to become friends at the end of the five-month period, than youngsters who were dissimilar in their physical activity. At the same time, youngsters who were already friends at the beginning of the study influenced the physical activity of their friends. Therefore, social network interventions might be preferable to individually oriented interventions.