There is much uncertainty in how climate change will impact the performance of urban trees. Climate niche modelling predicts that many urban tree species may become unsuitable in future climates, but this has rarely been tested in cities. Broad planting of diverse tree species in different cities provides the opportunity to test climate niche predictions. Here we investigated if the climate of origin of 14 urban tree species influenced tree growth, trait expression, and phenotypic plasticity. We determined climate niche limits for all species and measured canopy growth rates of individual trees from 2013 to 2021 in the two largest Australian cities: subtropical Sydney, and temperate Melbourne. Six functional traits including leaf water potential at turgor loss point (TLP), wood density (WD), leaf dry matter content (LDMC), specific leaf area (SLA), carbon isotope composition (δ13C) and Huber value (HV) were measured in both cities. Trees planted outside their climate niche limits had lower growth than trees planted inside their climate niche in the temperate but not subtropical city. Species with lower MAP of origin (i.e., drier) had faster canopy growth in both cities. Species with low MAP and high heat moisture index (HMI) at their origin had more negative TLP and greater WD, indicating species from drier environments maintain their high drought tolerance in cities. Trees planted in drier Melbourne had more negative TLP, higher WD and higher LDMC than in Sydney, demonstrating phenotypic plasticity in urban trees. Wetter origin species showed greater phenotypic plasticity in TLP, WD and δ13C. Canopy RGR was negatively related with δ13C reflecting a strong impact of stomatal behaviour on urban tree growth. Our study provides limited support that species climate niche limits reliably predict urban tree growth, so we caution against solely using climate niche matching and advocate for inclusion of functional traits when selecting urban tree species.