This experiment was conducted to clarify the effects of different soil moisture levels on the root distribution and root dry matter of upland-cultured paddy rice (cv. Myojoo, Kochihibiki), crossbred rice of paddy rice and upland rice (cv. Ishioka glutinous No.10, Mizuhatamochi), and upland rice (cv. Norin No. 12, Norin No. 21). In the first experiment (1977), the root distribution and dry matter was investigated on the six cultivars which were cultivated under irrigation and nonirrigation. The irrigation treatments were begun at July 18. The soil moisture tension of the two irrigation plots were at pF 2.5 and pF 2.0 at l0 cm depth, respectively. During July 24 to August 8, the soil moisture tension of nonirrigation plot was over pF 2.9 and in the range of pF 2.5 to 2.9, at 10 cm and 30 cm depths, respectively. Therefore, the soil moisture of nonirrigation plots reached to first permanent wilting point at l0 A.M., August 4. Top dry matter, root dry matter and root distribution were investigated at maturity. The investigation of the root distribution was practiced by excavating soil masses, which were sectioned by 7 cm increments up to 35 cm depth. The second experiment (1978) was set up in cylindrical plastic pots of 30 cm height and 25 cm inner diameter using four cultivars (Myojoo, Mizuhatamochi, Norin No. 12, Norin No. 21). Soil moisture treatments composed by two levels of high and low were begun at July 10. That is, high soil moisture plot was irrigated by each pots every day. While, low soil moisture plot was irrigated when every time the soil moisture became to first permanent wilting point. The treatments were continued for 58 days. The results showed that root distribution of the rice plants was effectively influenced by different soil moisture levels. The roots of rice plants distributed in the shallow layer under irrigation, while the roots distributed in the deeper layer under nonirrigation. Root distribution of upland-cultured paddy rice and crossbred rice of paddy rice and upland rice distributed in the shallow layer than that of upland rice. Also, root dry matter was influenced by different soil moisture levels. The root dry matter was increased in drier soil condition, where rice plants proliferated numerous branches to gather a small amount of soil water. The root dry matter of upland rice cultivar 'Norin No. 12' was increased conspicuously in surface layer under nonirrigation. It is suggested that high growth rate of root weights in water stressed rice plants was correlated with drought resistance, because high growth rate of root weight was obtained by upland rice which has drought tore-lance. R-T ratio was increased in drier treatments, because root dry matter was increased and top dry matter was decreased by water stress.