<p>Large amounts of C are allocated to tree roots, but little is known about the age and dynamics of their non-structural C (NSC). We measured bomb-radiocarbon (<sup>14</sup>C) in respired CO<sub>2</sub>, non-structural (mainly sugars), and structural (cellulose) C in roots. The steady decline of &#916;<sup>14</sup>C in atmospheric CO<sub>2</sub> since the 1960s indicates the mean time elapsed since the C in these pools was fixed. We measured coarse (>2 mm, mean 2.91 mm) and fine (<2 mm) roots from 12 German poplar trees sampled before and after girdling of 6 of the trees. All samples were taken in 2018, an exceptionally dry summer in Europe. The mean &#916;<sup>14</sup>C &#177;SD of root-respired CO<sub>2</sub> (4.1 &#177; 3.6 &#8240;) in June-July was equal to current atmospheric &#916;<sup>14</sup>CO<sub>2</sub> (1.2 &#8240;), irrespective of the mean age of root cellulose. During extended incubations, the &#916;<sup>14</sup>C of root-respired CO<sub>2</sub> increased to ~10 &#8240; 8 days after harvesting and up to 42 &#8240; 17 days after harvesting. The mean &#916;<sup>14</sup>C of soluble sugars in the roots was ~21&#160;&#8240;. In September-October, almost three months after girdling, roots from girdled trees respired CO<sub>2</sub> with &#916;<sup>14</sup>C of 7.9 &#177;&#160;6.6&#160;&#8240; vs. 2.3 &#177;&#160;6.1&#160;&#8240; in the ungirdled control trees. However, in both groups the respired CO<sub>2&#173;</sub>-&#916;<sup>14</sup>C correlated with cellulose-&#916;<sup>14</sup>C (R<sup>2</sup> = 0.37, 0.26 for girdled and control trees, respectively), suggesting that roots respired more stored C in the later growing season in this drought year, independent of treatment. The &#916;<sup>14</sup>C values of soluble sugars were correlated with the &#916;<sup>14</sup>C values of the cellulose (R<sup>2</sup>=0.83). On average, C in sugars was fixed more recently than cellulose, suggesting mixing of young C from other parts of the tree into the roots. Stem girdling did not affect the &#916;<sup>14</sup>C of soluble sugars. Average total sugar concentrations (sucrose+ glucose+ fructose) were ~42 mg g<sup>-1 </sup>and did not vary with sampling date, root class or treatment. Starch, measured only in September-October, was higher in coarse than in fine roots (12 vs. 3.8 mg g<sup>-1</sup>). Respiratory loss of C was higher in the fine roots (~4 mgC g<sup>-1</sup> day<sup>-1</sup>) than coarse roots (~2.4 mgC g<sup>-1</sup> day<sup>-1</sup>), with no effect of girdling or sampling month. When normalize (expressed per gram dry root material), the NSC reservoirs and C loss rates suggest C turnover rates are 2-fold higher in fine roots than in coarse roots. The extended incubations indicate that detached roots are able to quickly utilize stored NSC, as indicated by the sharp &#916;<sup>14</sup>CO<sub>2</sub> increase. In comparison, stem girdling had no measurable effect on respired CO<sub>2</sub>-&#916;<sup>14</sup>C, suggesting internal re-allocation of C from the lower stem base or large roots to smaller roots, and/or lower than expected metabolic consumption of C in reaction to girdling or because of the exceptional drought.</p>