Purpose The presumption in a cold sensitivity test (CST) used for cold injuries is that the skin temperature (T sk) observed reflects the return of blood flow to the extremity following a local cold challenge. We questioned this assumption. Methods Six non-cold injured participants undertook two CSTs in 30 °C air. The control (CON) CST involved 12 min gentle exercise prior to immersing the foot into 15 °C water for 2 min followed by 15 min of spontaneous rewarming. The occlusion (OCC) CST was the same except that blood flow to the foot was occluded during the rewarming period. These results were compared to CSTs from six individuals with non-freezing cold injury and moderate–severe cold sensitivity (CS) and a non-perfused human digit model (NPDM). Results Before immersion, great toe skin blood flow (SkBF) was similar in CON and OCC conditions [255 (107) laser Doppler units (LDU)] and was higher than CS [59 (52) LDU]. During rewarming, SkBF in CON returned to 104 % of the pre-immersion value and was higher than both OCC and CS. Great toe T sk before immersion was lower in CS [28.5 (2.1) °C] compared to CON [34.7 (0.4) °C], OCC [34.6 (0.9) °C] and NPDM [35.0 (0.4) °C]. During rewarming skin/surface temperature in OCC, CS and NPDM were similar and all lower than CON. Conclusions SkBF does contribute to the skin rewarming profile during a CST as a faster rate of rewarming was observed in CON compared to either OCC or NPDM. The lower T sk in CS may be due to a reduced basal SkBF.