ABSTRACT In a study of vibration-induced white finger (VWF) 39 chain saw operators, being representative of 365 lumberjacks according to a completed questionnaire, underwent clinical investigation. In a medical interview 13 of the 39 had VWF and 26 were without finger symptoms. In the subjects with VWF both hands were equally affected, the third and fourth finger most frequently; the thumbs were always normal. A cold provocation test, measuring the finger systolic blood pressure with a cuff technique during combined body cooling and finger cooling to 30, 15, and 6°C, was applied to the 39 chain saw operators and to 20 age-matched outdoor workers who had not used vibrating hand tools. The finger most often showing Raynaud's phenomenon was cold provoked, and a non-affected finger was used as a reference. The systolic pressure gradient measured at 30°C from the upper arm to the cooled finger was increased in the chain sawyers most severely affected by VWF compared with the control group and with the chain sawyers without VWF (p <0·02). Raynaud's phenomenon was verified by digital arterial closure (zero pressure in the finger) at 15 or 6°C in 12 of 13 chain sawyers having anamnestic and clinical evidence of VWF. Chain sawyers without VWF had an increased digital arterial response to cold at 15 and 6°C compared with the control group (p <0·05) and had a decreased cold response compared with chain sawyers with VWF (p <0·05). In chain sawyers without VWF 13 subjects with an exposure period of 10-23 years had an increased cold response at 6°C compared with 13 subjects with an exposure period of one to three years (p <0·05). Only one of 38 chain saws did not exceed the exposure limit proposed by the International Standardisation Organisation. The results of the present study indicate that vibrations induce hyperresponsiveness to cold in the exposed arteries, even in chain sawyers without finger symptoms. The exposure to vibration therefore should be reduced to eliminate this damage. The presented cold provocation test may be used to diagnose Raynaud's phenomenon and to disclose an abnormal digital arterial cold response before subjective symptoms have appeared.