Background: Schools are a key setting for large-scale primordial non-communicable disease prevention in young people, but little data on sustainability of impacts on cardiometabolic risk markers is available. Methods: Six and a half year follow-up of a natural experiment. In 2008, six public schools in the municipality of Svendborg (Denmark) augmented their curricular physical education (intervention) and four matched schools served as controls. At long term follow up in 2015 n = 312 participants aged 5-11 years had complete data (33% of children providing necessary baseline data). The intervention, that consisted of a trebling of weekly physical education lessons and courses provided to physical education teachers, was provided at intervention schools up until 6th grade. Participants attended 6th to 10th grade at follow-up. Differences in the homeostasis model assessment of insulin resistance, blood pressure, triglycerides, cholesterol ratios, cardiorespiratory fitness, waist-circumference, and a composite score of these, between participants attending intervention and control schools were analysed by mixed linear regression models. Differences in physical activity at follow-up was analysed cross-sectionally (no baseline available) in n = 495. Results: Compared to controls, children at intervention schools had a non-significant - 0.07 (- 0.32 to 0.18) standard deviations lower composite risk score 6.5 years after project initiation. Likewise, no statistically significant differences between intervention and control schools were found for any of the other outcomes (p-values ≥ 0.41). However, six of seven outcomes were in a direction favouring intervention schools. No statistically significant differences between intervention and control schools were observed for physical activity outcomes (p-values ≥ 0.13). Conclusions: An augmented physical activity program including 270 min of weekly physical education provided for three to seven years did not materialize in statistically significant differences in established risk markers in children from intervention compared to control schools. As the intervention was discontinued after 6th grade, the post-intervention effect of augmented physical education throughout adolescence is unknown. School-based physical activity programs may benefit from incorporating instruments for behaviour translation to leisure time in their intervention models to increase the probability of achieving public health relevance. Trial registration: ClinicalTrials.gov Identifier: NCT03510494.