The carbon isotopic composition (δ<sup>13</sup>C) of CO<sub>2</sub> efflux (δ<sup>13</sup>C<sub>efflux</sub>) from soil is generally interpreted to represent the actual isotopic composition of the respiratory source (δ<sup>13</sup>C<sub>Rs</sub>). However, soils contain a large CO<sub>2</sub> pool in air-filled pores. This pool receives CO<sub>2</sub> from belowground respiration and exchanges CO<sub>2</sub> with the atmosphere (via diffusion and advection) and the soil liquid phase (via dissolution). Natural or artificial modification of δ<sup>13</sup>C of atmospheric CO<sub>2</sub> (δ<sup>13</sup>C<sub>atm</sub>) or δ<sup>13</sup>C<sub>Rs</sub> causes isotopic disequilibria in the soil-atmosphere system. Such disequilibria generate divergence of δ<sup>13</sup>C<sub>efflux</sub> from δ<sup>13</sup>C<sub>Rs</sub> (termed "disequilibrium effect"). <br><br> Here, we use a soil CO<sub>2</sub> transport model and data from a <sup>13</sup>CO<sub>2</sub>/<sup>12</sup>CO<sub>2</sub> tracer experiment to quantify the disequilibrium between δ<sup>13</sup>C<sub>efflux</sub> and δ<sup>13</sup>C<sub>Rs</sub> in ecosystem respiration. The model accounted for diffusion of CO<sub>2</sub> in soil air, advection of soil air, dissolution of CO<sub>2</sub> in soil water, and belowground and aboveground respiration of both <sup>12</sup>CO<sub>2</sub> and <sup>13</sup>CO<sub>2</sub> isotopologues. The tracer data were obtained in a grassland ecosystem exposed to a δ<sup>13</sup>C<sub>atm</sub> of −46.9 ‰ during daytime for 2 weeks. Nighttime δ<sup>13</sup>C<sub>efflux</sub> from the ecosystem was estimated with three independent methods: a laboratory-based cuvette system, in-situ steady-state open chambers, and in-situ closed chambers. <br><br> Earlier work has shown that the δ<sup>13</sup>C<sub>efflux</sub> measurements of the laboratory-based and steady-state systems were consistent, and likely reflected δ<sup>13</sup>C<sub>Rs</sub>. Conversely, the δ<sup>13</sup>C<sub>efflux</sub> measured using the closed chamber technique differed from these by −11.2 ‰. Most of this disequilibrium effect (9.5 ‰) was predicted by the CO<sub>2</sub> transport model. Isotopic disequilibria in the soil-chamber system were introduced by changing δ<sup>13</sup>C<sub>atm</sub> in the chamber headspace at the onset of the measurements. When dissolution was excluded, the simulated disequilibrium effect was only 3.6 ‰. Dissolution delayed the isotopic equilibration between soil CO<sub>2</sub> and the atmosphere, as the storage capacity for labelled CO<sub>2</sub> in water-filled soil pores was 18 times that of soil air. <br><br> These mechanisms are potentially relevant for many studies of δ<sup>13</sup>C<sub>Rs</sub> in soils and ecosystems, including FACE experiments and chamber studies in natural conditions. Isotopic disequilibria in the soil-atmosphere system may result from temporal variation in δ<sup>13</sup>C<sub>Rs</sub> or diurnal changes in the mole fraction and δ<sup>13</sup>C of atmospheric CO<sub>2</sub>. Dissolution effects are most important under alkaline conditions.