The development of effective environmental policies is needed in order to meet regulatory standards and international legislation and agreements, both at national and supra-national level. The design, assessment and comparison of control strategies require the support of multi-scale integrated assessment modelling (IAM) systems. For instance, the development of ozone control strategies requires the understanding of the temporal and spatial multi-scale nonlinear processes governing tropospheric ozone levels and its complicated relationships with other atmospheric issues such as aerosols or acidification. IAM implies that science is being applied in the context of social and economic forces at work in society, so it has to be useful to generate both explanations and policy options. Regional (e.g. European) policies must rely on common, supra-national integrated assessment systems that necessarily imply a series of assumptions and simplifications that make them unsuitable for finer scales. More detailed and specific national systems are required to depict important phenomena that are out of reach of low-resolution models used at continental level. This is a challenging issue from the air quality modelling point of view, since model design, underlying hypotheses and input data must be consistent across the different systems and scales. This paper describes the methodology followed in the development of the air quality IAM system for the Iberian Peninsula (SIMCA) to deal with these multi-scale issues. Air quality simulation in SIMCA relies on the WRF-SMOKE-CMAQ system. The Weather Research and Forecasting (WRF) modelling system is used to develop the meteorological fields needed to reproduce the fate of air pollutant in the atmosphere. The model is initialized from global reanalysis and runs over a 48 km x 48 km spatial resolution European domain which provides boundary conditions for the national modelling domain with a spatial resolution of 3 km x 3 km. Meteorological outputs are then processed with the Meteorology-Chemistry Interface Processor (MCIP) which consistently interpolates and derives the variables required by the Community Multiscale Air Quality (CMAQ) system. This process involves data adaptation to a different staggered grid and potentially a different vertical-coordinate system. Likewise, the CMAQ system is run over the two nested domains using inputs from WRF and the Sparse Matrix Operator Kernel Emissions (SMOKE). This software provides an efficient and very flexible platform for emission processing. Although different formats and datasets are used for the emission inventories and projections feed to SMOKE, all of them (national inventory, EMEP inventory and Spain's Emission Projections, SEP) are based on official, consistent estimates, which, in turn, are used as inputs for European-scale integrated assessment modelling studies such as those carried out for the development of the European Air Quality Thematic Strategy on Air Quality. However, emission processing is adapted to the specific requirements and data availability of the different domains. This approach allows both, the incorporation of the best information available at different scales and the implementation of standards used in international modelling exercises (e.g. Eurodelta) and models such as RAINS/GAINS-EMEP. On the other hand, CMAQ outputs are perfectly suited for the inputs requirements of the air pollution effects modules of the national IAM system.