Soil sealing is one of the most pervasive forms of soil degradation that follows urbanization and, despite innovative pavements (i.e. pervious) are being installed in urban areas to mitigate it, there is little research on the effects of pervious pavements on soil water and carbon cycle and on the physiology of urban trees. The aim of this 4-year experiment was to assess the effects of three pavements, differing in permeability to water and gases, on some soil physical parameters, and on growth and physiology of newly planted Celtis australis and Fraxinus ornus. Treatments were: 1) impermeable pavement (asphalt on concrete sub-base); 2) permeable pavement (pavers on crushed rock sub-base); 3) porous design (porous pavement on crushed rock sub-base); 4) control (unpaved soil, kept free of weed by chemical control). Soil (temperature, moisture, oxygen content and CO2 efflux) and plant (above- and below-ground growth, leaf gas exchange, chlorophyll fluorescence, water relations) parameters were measured.All types of pavements altered the water cycle compared to unpaved soil plots, but this disturbance was less intense in porous pavements than in other soil cover types. Porous pavements allowed both higher infiltration and evaporation of water than both pavers and asphalt. Reduction of evaporative cooling from soil paved with permeable and impermeable pavements contributed to significant soil warming: at 20cm depth, soils under concrete pavers and asphalt were 4 and 5°C warmer than soil covered by porous pavements and unpaved soils, respectively. Thus, enhancing evaporation from paved soil by the use of porous pavements may contribute to mitigating urban heat islands. CO2 greatly accumulated under impermeable and permeable pavements, but not under porous pavements, which showed CO2 efflux rates similar to control. Soil oxygen slightly decreased only beneath asphalt.Growth of newly planted C. australis and F. ornus was little affected by pavement type. Tree transpiration rapidly depleted soil moisture compared to the not-planted scenario, but soil moisture did not fall below wilting point (particularly in the deeper soil layers, i.e. 40–50cm) in any treatment. While C. australis showed similar leaf gas exchange and water relations in all treatments, F. ornus showed a depression in CO2 assimilation and slight signs of stress of the photosynthetic apparatus when planted in soil covered with impermeable pavement.The effects of soil cover with different materials on tree growth and physiology were little, because newly planted trees have most of their roots still confined in the unpaved planting pit. Still, the reduction of soil sealing around the planting pit triggered the establishment of sensitive species such as ash. Further research is needed to assess the effects of different pavement types on established, larger trees.