The impact of physical inactivity on health is well accepted throughout the medical and health service community. However, the case has largely been established through epidemiological studies with adults. Substantial attention has been paid to the activity levels of children and adolescents, largely because of changing lifestyles that have threatened the opportunity to be active and also introduced attractive sedentary alternatives such as playing computer games. The research evidence that children have become less active to the point where it is seriously damaging their current and future health has been difficult to establish. This situation is due to difficulties in establishing sensitive health risk markers, and also with the assessment of the different elements of physical activity which in children and adolescents is a complex profile of social behaviours. Self report of activity is unreliable with young children, and objective measures are required that are cheap and effective with large samples and that are capable of measuring levels, volume and patterns of physical activity. Accelerometry in combination with diaries offers the best current solution for most activity–health relationships, and for informing intervention need and design.