Two allelopathic rice accessions, PI312777 and Allelopathyl, significantly suppressedthe growth of associated weeds in the field. Moreover, their weed-suppressing effectswere correlated with the cultivation patterns. The weed-suppressing effects of throwingand transplanting were more effective than that of direct seeding. Furthermore, theamounts of allelochemicals (resorcinols, flavones and hydroxamic acids) produced andreleased from two allelopathic rice accessions were much higher than that from a non-allelopathic rice variety Hua-Jing-Xian 1, and reached the maximum concentration at the6th leaf stage. Differences in the weed-suppressing effects of rice accessions appear toresult from the accessions producing and releasing different amounts of allelochemicalsin the field. Further research confirmed that in PI312777 plants, allelochemicals weresynthesized by the above-ground parts, and then secreted through the root tissues. Roottissues of PI312777 plants never produced the allelochemicals. Root exudates fromPI312777 could significantly inhibit the growth of E. crus-galli surrounding rice plantsin water culture. However, when activated carbon was added to the culture solution, whichcould absorb allelochemicals from root exudates, the growth of E. crus-galli was nolonger significantly inhibited. Weed-suppressing effects of rice accessions depended onallelopathy, cultivation patterns and other factors in rice fields, while allelopathywas one of important factors. Interestingly, the amounts of allelochemicals produced andreleased from allelopathic rice plants may be induced by the presence of E. crus-galli.This suggests that there is a possible chemical recognition between rice and E. crus-galli.