Physical activity (PA) is considered a cornerstone in weight control and public health guidelines recommend regular participation to prevent gains in adiposity. It may therefore come as a surprise that the cumulative evidence from observational studies to support this is not strong. A weakness of many published observational studies on this topic has been a reliance on a single baseline assessment of PA. Using only the baseline information on PA in a prospective study cause misclassification because of participants often change activity level during follow-up. In turn this causes regression dilution bias and decreases the precision of the estimate of an association between PA and adiposity. Furthermore, because gains in adiposity often are caused by a small average daily energy imbalance over many years, following individuals for longer periods of time is essential to characterize a relationship.