In a 4-yr field study, “weed suppressive” rice cultivars provided 30% greater control of barnyardgrass and sustained 44% less yield loss (relative to weed-free) compared to “nonsuppressive” tropical japonica rice cultivars. 13C analysis revealed that rice root mass predominated vertically and laterally within the soil profile of plots infested with barnyardgrass. Among all cultivars, rice roots accounted for 75 to 90% of the total root mass in samples, and this was most concentrated in the surface 5 cm of soil in the row. Barnyardgrass roots were most prevalent in the surface 5 cm between rows where they accounted for 30% of total root mass. Overall, barnyardgrass root mass was about twice as high in nonsuppressive rice compared to suppressive rice. Weed suppression by indica/tropical japonica rice crosses generally was intermediate between that of the other two rice groups. At the 0- to 5-cm depth, between-rows, barnyardgrass root mass was correlated negatively with rice height (r = −0.424), yield (r = −0.306), and weed control ratings (r = −0.524) in weedy plots. Control ratings in weedy plots also were negatively correlated with rice percent height reduction (r = −0.415) and % yield loss (r = −0.747) relative to weed-free plots, and with barnyardgrass root mass as a percent of total root mass (r = −0.612). Control ratings were positively correlated with rice yield under weed pressure (r = 0.429) but were correlated with rice root mass in-rows only (r = −0.322). Clearly, rice root mass could not have been the major cause of the differences in barnyardgrass control between cultivars.