Laboratory tests were conducted to investigate how much portland cement restrained a soil from frost heaving and thaw settlement when it had been added to the soil, using Fujinomori clay collected in Kyoto. They were prepared as follows : A slurry was made first by mixing thoroughly air-dried soil, portland cement and the proper amount of water necessary to make the slurry; next, it was let stand for 7 days to cure; then, a specimen was made in the shape of a column with two sizes : 5 cm in diameter and 10 cm in length; 10 cm in diameter and 4 cm in length. Let W_cs·denote the weight ratio in percentage of added cement to ovendried soil; V the initial volume of the specimen; ΔV the volume increment of the specimen after frost heaving; ζ the frost heave ratio in percentage (ζ=ΔV/V× 100); V_t the volume of the specimen after thawing; and ζ_s the thaw settlement ratio in percentage (ζ_s= (V_t-V) /V ×100). The value of e decreases to 1/81/4 times that of coment-free specimens when W_cs= 10%. The decrease in ζ is small when W_cs>10%. ζ_s is -2% for cement-free specimens; zero when W_cs =8%, and slightly positive when W_cs>8%. Theoretical consideration shows that a decrease in permeability of an unfrozen soil as a result of addition of cement leads to little decrease in ζ. Accordingly, it is suggested that a decrease in ζ comes about from a decrease in ability of the soil to absorb water because of the presence of cement in the soil. The tests reveal that mixing of soil with cement in the weight ratio of 10 : 1 restrains the mixture from frost heaving and thaw settlement.