To identify a set of important WBV predictors that could be used to develop a statistical instrument for exposure assessment in a large epidemiologic study, a total of 432 WBV measures were taken from a sample of 247 male drivers in Taipei City, Taiwan. In accordance with the ISO 2631-1 (1997) methods, we measured the frequency-weighted vertical acceleration (z-axis) over drivers' seat surface, under conditions representing different types of rides (vacant vs. short vs. long) assigned to random destinations. Mixed effect models were used to analyse the WBV data including repeated measures. For this group of urban taxi drivers regularly exposed to WBV of low intensity (mean = 0.31 ms( - 2), ranging from 0.17 to 0.55 ms( - 2) r.m.s.), our analyses indicated that average driving speed was the primary predictor (p < 0.0001). As average driving speed increased, measured vertical acceleration increased in a quadratic-linear manner (p < 0.0001). Other WBV predictors, after adjusting for the effects of other covariates, included automobile manufacturer (p = 0.02), engine size (p = 0.04), body weight (p = 0.002), age (p = 0.02), use of seat cushion (p = 0.03), and traffic period (p = 0.02). Our study suggests that a similar statistical approach could be employed in future studies to improve the quality and efficiency of WBV exposure assessment in professional drivers.