PURPOSE: Understanding seasonal variation in physical activity is important for informing public health surveillance and intervention design. The aim of the current study was to describe seasonal variation in children's objectively measured physical activity and sedentary time. METHODS: Data are from the UK Millennium Cohort Study. Participants were invited to wear an accelerometer for 7 d on five occasions between November 2008 and January 2010. Outcome variables were sedentary time (<100 counts per minute, min·d(-1)) and moderate to vigorous physical activity (MVPA) (>2241 counts per minute, min·d(-1)). The season was characterized using a categorical variable (spring, summer, autumn, or winter) and a continuous function of day of the year. Cross-classified linear regression models were used to estimate the association of each of these constructs with the outcome variables. Modification of the seasonal variation by sex, weight status, urban/rural location, parental income, and day of the week (weekday/weekend) was examined using interaction terms in regression models. RESULTS: At least one wave of valid accelerometer data was obtained from 704 participants (47% male; baseline age, 7.6 (0.3) yr). MVPA was lower in autumn and winter relative to spring, with the magnitude of this difference varying by weekday/weekend, sex, weight status, urban/rural location, and family income (P for interaction <0.05 in all cases). Total sedentary time was greater in autumn and winter compared with spring; the seasonal effect was stronger during the weekend than during the weekday (P for interaction <0.01). CONCLUSIONS: Lower levels of MVPA and elevated sedentary time support the implementation of intervention programs during autumn and winter. Evidence of greater seasonal variation in weekend behavior and among certain sociodemographic subgroups highlights targets for tailored intervention programs.