Cherry seedlings (Prunus avium) were grown from seed for two growing seasons in three ambient [CO2] (∼350 μmol mol−1) and three elevated [CO2] (ambient+∼350 μmol mol−1) open-top chambers, and in three outside blocks. A drying cycle was imposed in both the growing seasons to half the seedlings: days 69–115 in the first growing season, and in the second growing season days 212–251 on the same seedlings which had already experienced drought. Stomatal conductance was significantly reduced in elevated [CO2]-grown, unstressed seedlings in both the first and second growing seasons, but was not caused by a decrease in stomatal density. Droughted seedlings showed little or no reduction in stomatal conductance in response to elevated [CO2]. However, stomatal conductance was highly correlated with soil water status. Photosynthetic rate increased significantly in response to elevated [CO2] in both water regimes, leading to improvement in instantaneous transpiration efficiency over the whole duration of the experiment, but there was no relationship between instantaneous transpiration efficiency and long-term water use efficiency. The Amax was strongly reduced in the second growing season, but unaffected by [CO2] treatment. Although photosynthetic rate was not down-regulated, Rubisco activity was decreased by elevated [CO2], possibly because of the increased leaf carbon: nitrogen ratio which had occurred by the ends of the two growing seasons. Elevated [CO2] did not improve plant water relations (for example, bulk leaf – water potential, osmotic potentials at full and zero turgor, relative water content at zero turgor, bulk modulus of elasticity of the cell) and thus did not increase water-stress tolerance of cherry seedlings.