Abstract: Abstract: The freeze-thaw conditions affect soil shear strength and hence threaten engineering safety, soil slope stability, and soil loss in seasonal frozen-soil region. Shear strength was tested for 2 kinds of soils (sandy loam in Qinghai-Tibet Plateau (S1) and silt loam in Beijing region (S2)) under 3 freeze-thaw statuses (unfrozen soil, frozen soil, and thawing soil after frozen) and different initial water contents (8.0%-31.0% in mass fraction) by direct shear apparatus in the laboratory. The unfrozen soil specimen did not suffer freeze-thaw processing in the laboratory, the thawing soil specimen was sheared after frozen at -18℃ for 24 h, and the thawed soil specimen was done after frozen at -18℃ for 24 h and then thawed at 27℃ for 12 h. The shear velocity was set to 2.4 mm/min. Experimental results showed that the shear strength of the 2 types of soils increased approximately linearly with the increasing of normal stress, and the shear strength envelope of unsaturated thawing soil could be expressed by total stress Mohr-Coulomb failure criteria under tested normal stress, freeze-thaw status and soil water content conditions. The shear strengths of unfrozen and thawed soils were similar to each other and their difference was averagely 3.4%-3.7% after one freeze-thaw cycle, and they decreased with the increased initial soil water content for the 2 types of soils. The shear strength of soil sample S1 was 7.5%-9.7% greater than that of S2 under the unfrozen and thawed statuses. Under the thawing status, the shear strength of S1 increased with the initial soil water content but that of S2 decreased with it. The shear strength of the thawing S1 soil was smaller than that for the unfrozen and thawed soils at low soil water content (13.5%) but it was on the contrary when the water content was higher; the shear strength of the thawing S2 soil was smaller than that for the unfrozen and thawed soils under all the soil water contents. Under the unfrozen, thawed, and thawing statuses, the internal friction angle for both types of soils declined with the increase of soil water content. The internal friction angles under the unfrozen and thawed statues were similar to each other but they were significantly greater than that under the thawing status. The soil cohesion under the unfrozen or thawed statuses decreased with the increased soil water content and that under the thawed status was smaller than that under the unfrozen status for both types of soils. Under the thawing status, the soil cohesion of S1 increased significantly with the increased initial soil water content but that of S2 first increased and then decreased with it. The soil cohesion of both types of soils under the thawing status generally was significantly greater than that under the unfrozen and thawed statuses. The variation ranges of the internal friction angle for the 2 types of soils were similar to each other but their soil cohesion was significantly different under experimental conditions. The soil at the thawing status had a relatively low shear strength compared with that at the unfrozen or thawed statuses, and hence the thawing status should be adopted as the basic status for engineering design or soil loss prevention in seasonal frozen-soil regions.