Estimates of air pollution levels at fine spatiotemporal scale are lacking in Italy We combined satellite data with land use variables and ground-level PM10 measurements We estimated daily PM10 concentrations at a 1-km2 grid over Italy, 2006-2012 Our model displayed good cross-validation fitting (R2 = 0.65) and negligible bias Spatiotemporal predictions will allow estimation of short and long term health effects of PM10