The objectives of this paper were to review the literature on the responses of root systems to elevated CO2 in intact, native grassland ecosystems, and to present the results from a 2-yr study of root production and mortality in an intact calcareous grassland in Switzerland. Previous work in intact native grassland systems has revealed that significant stimulation of the size of root systems (biomass, length density or root number) is not a universal response to elevated CO2. Of the 12 studies reviewed, seven showed little or no change in root-system size under elevated CO2, while five showed marked increases (average increase 38%). Insufficient data are available on the effects of elevated CO2 on root production, mortality and life span to allow generalization about effects. The diversity of experimental techniques employed in these native grassland studies also makes generalization difficult. In the present study, root production and mortality were monitored in situ in a species-rich calcareous grassland community using minirhizotrons in order to test the hypothesis that an increase in these two measures would help explain the increase in net ecosystem CO2 uptake (net ecosystem exchange) previously observed under elevated CO2 at this site (600 vs 350 μl CO2 l−1; eight 1.2-m2 experimental plots per CO2 level using the screen-aided CO2 control method). However, results from the first 2 yr showed no difference in overall root production or mortality in the top 18 cm of soil, where 80–90% of the roots occur. Elevated CO2 was associated with an upward shift in root length density: under elevated CO2 a greater proportion of roots were found in the upper 0–6-cm soil layer, and a lower proportion of roots in the lower 12–18 cm, than under ambient CO2. Elevated CO2 was also associated with an increase in root survival probability (RSP; e.g. for roots still alive 280 d after they were produced under ambient CO2, RSP = 0.30; elevated CO2, RSP = 0.56) and an increase (48%) in median root life span in the deepest (12–18 cm) soil layer. The factors driving changes in root distribution and longevity with depth under elevated CO2 were not clear, but might have been related to increases in soil moisture under elevated CO2 interacting with vertical patterns in soil temperatures. Thus extra CO2 taken up in this grassland ecosystem during the growing season under elevated CO2 could not be explained by changes in root production and mortality. However, C and nutrient cycling might be shifted closer to the soil surface, which could potentially have a substantial effect on the activities of soil heterotrophic organisms as CO2 levels rise.