During microbial breakdown of leaf litter a fraction of the C lost by the litter is not released to the atmosphere as CO(2) but remains in the soil as microbial byproducts. The amount of this fraction and the factors influencing its size are not yet clearly known. We performed a laboratory experiment to quantify the flow of C from decaying litter into the soil, by means of stable C isotopes, and tested its dependence on litter chemical properties. Three sets of (13)C-depleted leaf litter (Liquidambar styraciflua L., Cercis canadensis L. and Pinus taeda L.) were incubated in the laboratory in jars containing (13)C-enriched soil (i.e. formed C4 vegetation). Four jars containing soil only were used as a control. Litter chemical properties were measured using thermogravimetry (Tg) and pyrolysis-gas chromatography/mass spectrometry-combustion interface-isotope ratio mass spectrometry (Py-GC/MS-C-IRMS). The respiration rates and the delta(13)C of the respired CO(2) were measured at regular intervals. After 8 months of incubation, soils incubated with both L. styraciflua and C. canadensis showed a significant change in delta(13)C (delta(13)C(final) = -20.2 +/- 0.4 per thousand and -19.5 +/- 0.5 per thousand, respectively) with respect to the initial value (delta(13)C(initial) = -17.7 +/- 0.3 per thousand); the same did not hold for soil incubated with P. taeda (delta(13)C(final:)-18.1 +/- 0.5 per thousand). The percentages of litter-derived C in soil over the total C loss were not statistically different from one litter species to another. This suggests that there is no dependence of the percentage of C input into the soil (over the total C loss) on litter quality and that the fractional loss of leaf litter C is dependent only on the microbial assimilation efficiency. The percentage of litter-derived C in soil was estimated to be 13 +/- 3% of total C loss.