IntroductionSleep curtailment is associated with obesity in children, but few studies have investigated this relationship in a longitudinal sample of adolescents. The aim of the present study was to examine the longitudinal association between weekday time in bed (TIB) at age 10–13 and overweight at age 16–19.MethodsAdolescents and their parents (N = 3025 families), participating in a longitudinal population‐based study, completed questionnaires assessing habitual bedtime and wake time on weekdays, weight and height, socioeconomic status (SES), internalizing mental health problems and disturbed eating. Two surveys were administered with a 6‐year interval (T1 and T2). A one‐way analysis of covariance (ANCOVA) was performed examining the association between TIB and weight category 6 years later, with SES, internalizing problems and disturbed eating at baseline entered as covariates. Hierarchical and logistic regression analyses were used to assess TIB at age 10–13 years to as a predictor of body mass index (BMI) standardized deviation scores (SDS) and overweight status at age 16–19 adjusting for the same confounders and baseline BMI.ResultsA linear inverse relationship between TIB at age 10–13 and BMI category at age 16–19 was demonstrated by the ANCOVA, p < 0.001. Shorter TIB was related to higher weight, but the effect size was small (partial eta squared = 0.01). When adjusting for the included baseline confounders in the hierarchical regression model TIB significantly predicted later BMI SDS (β = −0.039, p = 0.02). The adjusted logistic regression model showed that for each hour reduction of TIB at T1 the odds of being overweight/obese at T2 increased with a factor of 1.6.ConclusionShorter TIB was found to be a significant, yet modest, independent predictor of later weight gain in adolescence. The findings implicate that establishing healthy sleep habits should be addressed in prevention and treatment strategies for adolescent obesity.