We developed an integrated approach coupling a chemical transport model (CTM) with machine learning (ML) techniques to produce high spatial resolution NO 2 and O 3 daily concentration fields over Italy. Three years (2013–2015) simulations, at a spatial resolution of 5 km, performed by the Flexible Air quality Regional Model (FARM) were used as predictors, together with other spatial-temporal data, such as population, land-use, surface greenness and road networks, by a ML Random Forest (ML-RF) algorithm to produce daily concentrations at higher resolution (1 km) over the national territory. The evaluation of the adopted integrated approach was based on NO 2 and O 3 observations available from 530 and 293 monitoring stations across Italy, respectively. A good performance for NO 2 and excellent results for O 3 were obtained from the application of the CTM; as for NO 2 , the levels at urban traffic stations were not captured by the simulations due to the adopted horizontal resolution and related emissions uncertainties. Performance improvements were achieved with ML-RF predictions, reducing NO 2 underestimation (near zero fractional bias results) and better capturing spatial contrasts. The results obtained in this work were used to support the national exposure assessment and environmental epidemiology studies planned in the BEEP (Big data in Environmental and occupational Epidemiology) project and confirm the potential of machine learning methods to adequately predict air pollutant levels at high spatial and temporal resolutions. Graphical abstract