Total and basal respiration (R_t and R_b , respectively) of intact and undisturbed roots of one-year-old Douglas fir seedlings, Pseudotsuga menziesii var. glauca [Beissn] Franco, were measured at experimentally varied soil carbon dioxide concentrations ([CO₂ ]). Use of specially designed root boxes and a CO₂ gas-flow compensating system designed around an infrared gas analyzer (IRGA) allowed controlled delivery of CO₂ to roots and simultaneous measurements of CO₂ released by roots. Root respiration rate responded to each inlet [CO₂ ], independent of whether the previous concentration had been higher or lower, within two to three hours (paired t test = 0.041, P= 0.622, and n= 13). Total and basal respiration rates decreased exponentially as soil [CO₂ ] rose from 130 ppm, well below atmospheric [CO₂ ], to 7015 ppm, a concentration not uncommon in field soils. Analyses of variance (ANOVA) showed that the effects of soil [CO₂ ] on rates of total and basal root respiration were statistically significant. Root respiration rates decreased by 4 to 5 nmol CO₂ g^-1 dry weight of roots s^-1 for every doubling of [CO₂ ] according to the following equations: ln(R₁ ) (nmol CO₂ g^-1 s^-1 ) = 5.24-0.30*ln[CO₂ ] with r= 0.78, P < 0.0001, and n= 70; and ln(R_b ) (nmol CO₂ g^-1 s^-1 ) = 6.29-0.52 * ln[CO₂ ] with r= 0.82, P < 0.0001, and n= 35. The sensitivity of root respiration to [CO₂ ] suggests that some previous laboratory measurements of root respiration at atmospheric [CO₂ ], which is 3 to 10-fold lower than [CO₂ ] in field soils, overestimated root respiration in the field. Further, the potential importance of soil [CO₂ ] indicates that it should be accounted for in models of below-ground carbon budgets.