Are mature forest trees carbon limited at current CO2 concentrations? Will ‘mid-life’, 35 m tall deciduous trees grow faster in a CO2-enriched atmosphere? To answer these questions we exposed ca. 100-year-old temperate forest trees at the Swiss Canopy Crane site near Basel, Switzerland to a ca. 540 ppm CO2 atmosphere using web-FACE technology. Here, we report growth responses to elevated CO2 for 11 tall trees (compared with 32 controls) of five species during the initial four treatment years. Tested across all trees, there was no CO2 effect on stem basal area (BA) increment (neither when tested per year nor cumulatively for 4 years). In fact, the 4th year means were almost identical for the two groups. Stem growth data were standardized by pretreatment growth (5 years) in order to account for a priori individual differences in vigor. Although this experiment was not designed to test species specific effects, one species, the common European beech, Fagus sylvatica, showed a significant growth enhancement in the first year, which reoccurred during a centennial drought in the third year. None of the other dominant species (Quercus petraea, Carpinus betulus) showed a growth response to CO2 in any of the 4 years or for all years together. The inclusion or exclusion of single individuals of Prunus avium and Tilia platyphyllos did not change the picture. In elevated CO2, lateral branching in terminal shoots was higher in Fagus in 2002, when shoots developed from buds that were formed during the first season of CO2 enrichment (2001), but there was no effect in later years and no change in lateral branching in any of the other species. In Quercus, there was a steady stimulation of leading shoot length in high-CO2 trees. Phenological variables (bud break, leaf fall, leaf duration) were highly species specific and were not affected by elevated CO2 in any consistent way. Our 4-year data set reflects a very dynamic and species-specific response of tree growth to a step change in CO2 supply. Stem growth after 4 years of exposure does not support the notion that mature forest trees will accrete wood biomass at faster rates in a future CO2-enriched atmosphere.