The reliability and validity of two tests (cold water and reactive hyperaemia) designed to confirm a patient's history of vibration induced white finger were studied. The cold water test is a measure of digital rewarming after hand immersion in cold water. Reactive hyperaemia consists of measuring digital rewarming after cold water immersion plus temporary ischaemia imposed on the hand. For ten weeks, ten healthy male volunteers were submitted once a week to both tests to study their reliability. The results showed a strong inter and intraindividual scattering. The mean value for the whole group, however, did not differ significantly from one week to the next. Fifty two subjects exposed to hand/arm vibration were submitted to both tests to estimate their validity. They were classified, according to their medical history, into three groups: A = no symptoms, B = tingling or numbess, or both, C = Raynaud's phenomenon. Both tests agreed with the clinical staging. For reactive hyperaemia, however, the differences between the groups were statistically significant only when the test was performed at 10 degrees C. These tests are more useful to study a group than an individual case. Time has no significant effect on the mean result of a group.