Heat waves that trigger severe droughts are predicted to increase globally; however, we lack an understanding of how trees respond to the combined change of extreme temperatures and water availability. Here, we studied the impacts of two consecutive heat waves as well as post-stress recovery in young Pseudotsuga menziesii (Mirb.) Franco (Douglas-fir) and Robinia pseudoacacia L. (black locust) growing under controlled conditions. Responses were compared under water supply close to the long-term average and under reduced irrigation to represent drought. Exposure to high temperatures (+10 °C above ambient) and vapour pressure deficit strongly affected the trees in terms of water relations, photosynthesis and growth. Douglas-fir used water resources conservatively, and transpiration decreased in response to mild soil water limitation. In black locust, heat stress led to pronounced tree water deficits (stem diameter shrinkage), accompanied by leaf shedding to alleviate stress on the hydraulic system. The importance of water availability during the heat waves became further apparent by a concurrent decline in photosynthesis and stomatal conductance with increasing leaf temperatures in both species, reaching the lowest rates in the heat–drought treatments. Stress severity determined both the speed and the amount of recovery. Upon release of stress, photosynthesis recovered rapidly in drought-treated black locust, while it remained below control rates in heat (t = −2.4, P < 0.05) and heat–drought stressed trees (t = 2.96, P < 0.05). In Douglas-fir, photosynthesis recovered quickly, while water-use efficiency increased in heat–drought trees because stomatal conductance remained reduced (t = −2.92, P < 0.05). Moreover, Douglas-fir was able to compensate for stem-growth reductions following heat (−40%) and heat–drought stress (−68%), but most likely at the expense of storage and other growth processes. Our results highlight the importance of studying heat waves alongside changes in water availability. They further suggest that we should look beyond the actual stress event to identify lagged effects and acclimation processes that may determine tree resilience in the long term.