Consumption-based material footprints calculated with multi-regional input–output (mrIO) analysis are influenced by the sectoral, spatial and material aggregations used in the mrIO tables, and lack of disaggregation can be a source of uncertainty. This study investigated the effect of the resolution of mrIO databases on consumption-based material footprints. The effect of aggregation was investigated by constructing input–output tables with different spatial, product and material category resolutions and comparing the calculated material footprints. Our results indicate that the material footprints of countries calculated using the different spatial and product aggregations are in general in the order of a few percent, with outliers in the order of 25% difference. The use of IO models with a low product category resolution (e.g. 60 product categories) to calculate the embodied material use of individual products will likely result in inaccurate estimations of the total embodied material for some product categories. Aggregating the original 46 material categories into 16 categories changes the calculated material footprint of countries by about 30%. This result strongly suggests that the material data used to create the extensions for the IO framework should be collected at the highest resolution that is practically feasible. •Effect of aggregation on material footprint versus carbon footprint•Material footprints of countries and products.•Aggregation at different stage of the creation of multi-regional input-output table.