A detailed understanding of the processes that contribute to the δ13C value of respired CO2 is necessary to make links between the isotopic signature of CO2 efflux from the soil surface and various sources within the soil profile. We used density fractionation to divide soils from two forested sites that are a part of an ongoing detrital manipulation experiment (the Detrital Input and Removal Treatments, or DIRT project) into two soil organic matter pools, each of which contributes differently to total soil CO2 efflux. In both sites, distinct biological pools resulted from density fractionation; however, our results do not always support the concept that the light fraction is readily decomposable whereas the heavy fraction is recalcitrant. In a laboratory incubation following density fractionation we found that cumulative respiration over the course of the incubation period was greater from the light fraction than from the heavy fraction for the deciduous site, while the opposite was true for the coniferous site.Use of stable isotopes yielded insight as to the nature of the density fractions, with the heavy fraction solids from both forests isotopically enriched relative to those of the light fraction. The isotopic signature of respired CO2, however, was more complicated. During incubation of the fractions there was an initial isotopic depletion of the respired CO2 compared to the substrate for both soil fractions from both forests. Over time for both fractions of both soils the respired δ13C reflected more closely the initial substrate value; however, the transition from depleted to enriched respiration relative to substrate occurs at a different stage of decomposition depending on site and substrate recalcitrance. We found a relationship between cumulative respiration during the incubation period and the duration of the transition from isotopically depleted to enriched respiration in the coniferous site but not the deciduous site. Our results suggest that a shift in microbial community or to dead microbial biomass as a substrate could be responsible for the transition in the isotopic signature of respired CO2 during decomposition. It is likely that a combination of organic matter quality and isotopic discrimination by microbes, in addition to differences in microbial community composition, contribute to the isotopic signature of different organic matter fractions. It is apparent that respired δ13CO2 cannot be assumed to be a direct representation of the substrate δ13C. Detailed knowledge of the soil characteristics at a particular site is necessary to interpret relationships between the isotopic values of a substrate and respired CO2.