The aims of this study were (1) to measure frequency-weighted vibration acceleration and (2) to study the effects of introducing a vibration-proof impact wrench on VWF in workers. The subject pool was 383 male workers who were regularly using an impact wrench and taking special medical examinations for vibration syndrome in a factory from 1982 to 1999. The prevalence of workers with VWF increased gradually after 1982, reached a peak value (4.8%) in 1986, gradually decreased after 1987, and disappeared in 1994. Sixteen subjects who had had VWF at least one time during the observation period were selected for this study. The stages of VWF were at stage I on the Stockholm Workshop scale in all subjects. After the vibration-proof impact wrench was introduced in 1986, the vibration acceleration of the impact wrench measured on the handle decreased from 8.6-11.1 m/s2 to 5.1-7.1 m/s2. The actual time per day that subjects were assumed to use the impact wrench was 108 minutes. The subjects actually used an impact wrench more than the occupational exposure limit allowed. However, VWF disappeared after the introduction of a vibration-proof impact wrench. This might have resulted from the combined effect of introducing the vibration-proof impact wrench and certain countermeasures that were taken against cold working environments.