Predicted and observed prevalences and latency periods of vibration-induced white finger (VWF) were examined among workers exposed to hand-arm vibration. The different physical characteristics of vibration--spectra and impulsiveness--were measured. The following groups of workers were included in the study: forest workers (n = 199), pedestal grinders (n = 12), stone workers (n = 16), shipyard workers (n = 171), and platers (n = 5). The exposure to vibration was measured according to the ISO 5349 method. The impulsiveness of vibration was defined as the difference between peak levels and RMS levels. A good agreement was observed between the predicted and observed data for prevalence and latency of VWF in the forest workers. For the tools with high impulsiveness used in grinding, stone works, and shipyard assembly hall, the results were nonconfirmative; and there was a poor correlation between vibration and VWF. The ISO 5349 standard does not consider the high peak values of the vibration signal which may comprise high-frequency components and cause short transients in the underlying tissue of the worker's hand. These characteristics in vibration may be hazardous in the genesis of VWF and cannot be predicted when measuring vibration by the present standard method.