Root growth and respiration in elevated CO2 (700 μmol mol-1 ) was studied in three tree species, Fraxinus excelsior L., Quercus petraea. L. and Pinus sylvestris L. grown in open-top chambers (OTCs) during a long-term exposure (20 months), during which root systems were allowed to develop without restriction imposed by pots. Root growth, measured as root length using root in-growth bags was increased significantly in trees exposed to elevated CO2 , although the magnitude of the response differed considerably between species and with time of sampling, the greatest effect observed after 6 months in ash (ratio of elevated: ambient, e∶a; 3·40) and the smallest effect observed in oak (e∶a; 1·95). This was accompanied by changes in specific root length, with a significant decrease in all species after 6 months, suggesting that root diameter or root density were increased in elevated CO2 . Increases in root length might have resulted from an acceleration in root cell expansion, since epidermal cell size was significantly increased in the zone of elongation in ash root tips (P2 were observed for root carbohydrates, with significant increases in soluble sugars for all species (P2 (P2 increased the total root d. wt for whole trees of all three species after 8 months of exposure, although the magnitude of this effect, in contrast to the root in-growth study, was greatest in Scots pine and smallest in ash. No significant effect of elevated CO2 was observed on the root∶shoot ratio. Further detailed analysis of whole root systems after 20 months confirmed that species differences in root responses to elevated CO2 were apparent, with increased coarse and fine root production in elevated CO2 for Scots pine and ash respectively. Lateral root number was increased in elevated CO2 for all species, as was mean root diameter. Root respiration rates were significantly reduced in elevated CO2 for all three species. These results provide firm evidence that exposure of trees to future CO2 concentrations will have large effects on root system development, growth, carbohydrate status and respiration. The magnitude and direction of such effects will differ, depending on species. The consequences of such responses for the three species studied are discussed.