Aims: To evaluate an epidemiological approach to a national noise hazard surveillance strategy, and report current exposure levels in the Danish workplace. Methods: A study base of 840 companies in 10 selected high risk industries in the largest county in Denmark was identified from a national register. Noise exposure was measured among manual workers recruited from a random sample of workplaces in each industry. For reference, financial companies and a sample of residents were investigated according to the same protocol. The A-weighted equivalent sound level (L Aeq ) for a full shift was measured by portable dosimeters worn by 830 workers employed at 91 workplaces (67% of 136 eligible companies). Results: The epidemiological design proved feasible and established a baseline for future noise surveillance. Substantial resources were needed to motivate workplaces to enlist and the final participation rate was less than optimal (66.9%). The L Aeq (8) values in the selected industries were highly elevated (mean 83.7 dB(A) (95% CI 83.3 to 84.1) in comparison with residents and office workers (mean 69.9 dB(A), 95% CI 68.8 to 71.0). Some 50% of the workers were exposed to more than 85 dB(A) and some 20% to more than 90 dB(A) in several industries. Conclusion: Noise levels in Danish high risk industries remain high. A substantial proportion of workers are exposed to noise levels above the current threshold limit of 85 dB(A). Ongoing surveillance of noise exposure using full shift dosimetry of workers in random samples of workplaces most at risk to high noise levels may help reinforce preventive measures. Such a programme would benefit from compulsory workplace participation.