Traffic estimation is an important area in official statistics and is used by policymakers in their decision-making process for regional planning. This paper describes a framework that relies on combining large amounts of data from traffic loop sensors with administrative data of the entire population of the Netherlands to estimate traffic intensities during rush-hour for all major roads in the Netherlands. Multiple calibration models are developed and compared, and although the models do suggest that traffic counts are sometimes over- or underestimated in some regions, the models overall perform well.