Assessing belowground plant competition is complex because it is very difficult to separate weed and crop roots from each other by physical methods. Alternative techniques for separating crop and weed roots from each other are needed. This article introduces a stable isotope method that can quantify the amounts of roots of rice and barnyardgrass intermixed in flooded field soils. It relies on the biological principle that rice, a C3 (photosynthetic pathway) species, discriminates more effectively than barnyardgrass, a C4 species, against a relatively rare isotopic form (13C) of CO2. This results in different 13C: 12C isotope ratios (expressed as δ13C) in root tissues of the two species. δ13C values for monoculture barnyardgrass and rice grown in a standard flood-irrigated system were highly stable over 4 crop-years, averaging −13.12 ± 0.80 (SD) and −28.5 ± 0.11 (SD)‰, respectively, based on analysis by an isotope ratio mass spectrometer. Standard concentration curves relating measured δ13C values to set proportions of rice:barnyardgrass root biomass were described by linear regressions, typically with r2 values of 0.96 or greater. Quantities of intermixed rice and barnyardgrass roots sampled 0 to 5 cm deep from soil between rice rows were estimated by extrapolation from standard curves based on δ13C values. About 50% more barnyardgrass root tissue was detected in plots of Lemont long-grain rice than in weed-suppressive PI 312777 indica rice, demonstrating the feasibility of using this stable carbon isotope method in flooded rice systems.