The objective of this study was to compare objective and subjective methods of collecting exposure time data for hand arm vibration (HAV) and whole-body vibration (WBV), and to evaluate the impact of inaccurate exposure times' on the calculation of the average vibration exposure over an 8 h working day A(8).The study was carried out in the engineering services and maintenance departments of a construction and property management company. Worker exposure time data was collected using three methods, questionnaire surveys, daily worker interviews and 8 h direct workplace observations. Vibration magnitudes (m/s(2)) were measured for a range of hand tools and vehicles, and daily vibration exposure estimates A(8) were calculated using exposure times observed, reported in interview and self reported in the questionnaire.Results from the study showed that self-reported exposure time estimates from the questionnaire survey were a factor of 9.0 (median value) times greater for HAV and a factor of 6.0 (median value) times greater for WBV when compared with direct observation estimates. Exposure times reported in interview were higher, than those observed, but more reliable than those self reported in the questionnaire: a factor of 2.1 (median value) times greater for HAV and a factor of 1.4 (median value) times greater for WBV. A(8) values calculated using questionnaire exposure times were up to 66% and 75% greater for sources of HAV and WBV respectively when compared to A(8) values calculated using observed exposure times.For the purposes of carrying out a reliable risk assessment, results from this study indicate that direct measurements of worker exposure time are not recommended over questionnaires especially where work is highly variable for example in construction and property management. Worker interviews or direct workplace observation methods were found to be reliable alternative methods for collecting exposure time. (C) 2010 Elsevier Ltd. All rights reserved.