The growing share of Europe's population living in cities makes urban climate change impact assessment and adaptation a critical issue. The urban environment is characterized by its sensitivity to small-scale meteorological, hydrological and environmental processes. These are generally not fully described in climate models, largely because of the models' insufficient spatial resolution. The Urban SIS climate service offers historical and future simulated data downscaled to 1 km × 1 km resolution over selected European metropolitan areas. The downscaled data are subsequently used as input to air quality and hydrological impact models, all made available to users as Essential Climate Variables and Sectoral Impact Indicators through a web portal. This paper presents the Urban SIS climate service and demonstrates its functionality in a case study in Stockholm city, Sweden. Good model performance was attained for intra-city temperature gradients and small-scale precipitation extremes. Less positive results were obtained for large-scale precipitation and hydrology, mainly due to an insufficient domain size in the meteorological and climate modelling, in turn related to the substantial computational requirements. An uncertainty classification approach was developed to aid the interpretation and user application of the data. We hope our lessons learnt will support future efforts in this direction.