This study examined the effect of cold water finger immersion at various water temperature on cold-induced vasodilation (CIVD) and its reproducibility to the cold stress. Ten healthy collegiate men ($21.4{pm}2.5$ yrs, $175.8{pm}4.1$ cm, $69.6{pm}7.6$ kg, $11.2{pm}3.7$ %fat) underwent two tests. At the first test (1ST), subjects immersed their middle fingers at $43^{circ}C$ water for 5 min followed by a resting at an ambient air for 25 min. Then they immersed the finger at one of the five water temperatures (Tw: 5, 8, 11, 14, or $17^{circ}C$) at random order for 20 min. Once a testing at one Tw was completed, they immediately repeated the testing procedure for another Tw. The second test(2ND) was performed within a week after 1ST with having an identical procedure of 1ST except the order of Tw. During the test, rectal temperature, finger temperature from middle finger nail bed, and heart rate were measured every six second. In conclusion, maximal finger temperature(Tfmax), and Tfmax minus Tw was highly reproducible in this experiment. Minimal finger temperature (Tfmin) and Tfmax were higher as Tw decreased. And Tfdiff was higher as the colds tress decreased. No differences were found in time variables of temperature responses.