In contrast to international trade, it is still difficult to find regionaltrade statistics within a nation. Given that the gravity model con-tinues to be very popular, we test two gravity-RAS approaches usedfor interregional trade estimation: a standard one and an extendedversion, which additionally estimates intra-regional flows. We assessthe accuracy with the help of two measures and for different sec-toral aggregation levels. For that, we use the survey-based 2005interregional input–output table of Japan as a benchmark. Resultsshow high overall accuracy levels for the standard approach, bet-ter than when using international data, albeit with heterogeneouserrors for sectors and regions. We further find that the results of amultiregional input-output model are highly sensitive to the tradeestimation approach and that errors slightly increase for increasingsectoral disaggregation levels. Results from an experiment based ona random number generator show how RAS influences results.