Gelifluction, thixotropy and yield stress were measured from < 5 g soil samples taken from Signy, Alectoria, Greenwich, Wiencke and Livingston islands in the maritime Antarctic using a temperature-controlled rotational rheometer. The small sample size that this method permitted is compatible with sampling from sensitive sample locations. An oscillating 10 Pa shear stress was applied to samples at -0.5 kPa water potential. Two freeze-thaw cycles had temperature ramps from 5°C to -10°C over 2 h, followed by -10°C to 5°C over 2 h and finally at 5°C for 1 h. At freezing onset, the shear modulus, G, dropped to 4–50 per cent of thawed G, with no differences between locations. At thawing onset, G dropped to 8–32 per cent of thawed G, with significant differences between locations (P < 0.001). Thixotropy was then measured by applying a 2 kPa oscillating shear stress for 10 min, followed by relaxation at 10 Pa for 2 h. The increased shear stress caused G to drop to less than 8 per cent of the pre-stressed value, with no difference between locations. After 0.1 and 2 h, G was 18–65 per cent and 31–82 per cent of the pre-stressed value, respectively. A shear ramp determined yield stresses ranging from 494–2217 Pa. These findings demonstrate the potential risk of more frequent freeze-thaw cycles or the occurrence of thawed soil to the stability of polar soils. Gelifluction through more frequent freeze-thaw cycles could result in increased slope movement, whereas thixotropy caused by trampling of thawed soils could exacerbate mechanical damage of surface soils.