Water availability is widely recognized as being an essential factor for tree survival, growth and for maximizing their ability in mitigating urban heat islands (UHI) through evapotranspiration. In urban areas, where the ground surface is highly impervious and the trees are not regularly irrigated, the reduction of precipitation infiltration into soils may increase water stress for trees (Whitlow et al., 1992; Clark and Kjelgren, 1990). It is also generally predicted that trees in urban sites have higher water losses than trees in natural forests due to increased evapotranspiration demand (McCarthy and Pataki, 2010) and that their average lifespan is shorter (Sæbø et al. 2003). Tree lifetime is constrained by physiological processes. When trees are old or when their physiological functioning are limited, trees stop growing and use their stored reserves. When these reserves are drained, trees decline and die (Downer, 2011). There is currently insufficient data to generalize the physiological responses of trees to the complex urban environment, where both climatic and management factors are entangled. Especially, little information is available on the effect of water stress on tree health, and its consequence on ecosystem services such as UHI mitigation. Furthermore, on-going climatic changes make it all the more necessary to anticipate the potential trajectories of urban trees, in terms of (i) risk assessment of tree survival, and (ii) their potential ability for cooling services. In this context, a retrospective approach of the long-term relations of trees to urban climate can thus provide a way to both enhance our understanding of current urban tree hydric state and gain insights on future levels of water stress levels under new climates. It is well known that there is a close relationship between tree growth and climate. Indeed, the size and the state of tree-rings are affected by the yearly sequences of favourable and unfavourable climates (Fritts, 1976). In turn, climate phenomena can be identified and reconstructed through ringwidth sequences (Hughes, 2002). Thus, dendrochronology can be used as informative tool to understand the long-term influence of past climate on urban trees growth. Consequently, understanding the past trajectory of tree growth under past climates can provide insights on their answer to future climate projections. Since tree cooling potential is tightly linked to water availability, negative feedback of water stress to tree cooling potential can be expected. The aim of this study is to investigate (i) to reconstruct the past growth dynamics of trees in urban environments using dendrochronology methods and principles and (ii) to compare the past growth of trees grown in contrasted environments (namely: streets, park and arboretum). These results can help determining urban environmental factors that impact urban trees growth and health the most. They can also be used as preliminary arguments