Objectives The LIMS databank stores all occupational exposure measurements collected by public health teams in Quebec. The IMIS databank contains exposure measurements collected by the U.S. Occupational Safety and Health Administration inspectors to verify compliance. We investigated differences/similarities between both databanks.

Methods Personal exposure measurements from 1994–2011 were abstracted, and industry from both databanks was recoded according to the Canadian Standard Industrial Classification. Logistic regression was used to explore differences between exposure levels in both databanks for 27 chemicals in common industries, taking into account the type of exposure (short-term or long-term), year, and industry. Ratios of the predicted odds of exposure above the threshold limit value (O_ER) for IMIS compared to LIMS were calculated for each chemical for years 1997 and 2008.

Results Our analysis was based on 64,938 LIMS and 53,078 IMIS measurements. Exposure levels were significantly lower in IMIS compared to LIMS for metals (O_ER estimated in 1997: 0.43 across agents, 95% confidence interval (CI): 0.30–0.62; O_ER 2008: 0.57, 95% CI: 0.42–0.77), and they became similar in recent years for solvents (O_ER 1997: 1.47, 95% CI: 0.91–2.38; O_ER 2008: 0.99, 95% CI: 0.58–1.69). Short-term exposure levels were on average 3 times higher than long-term ones across the two databanks. Results were unchanged when industry from both databanks was recoded according to the U.S Standard Industrial Classification or the North American Industry Classification System.

Conclusions Differences between exposure levels in the two databanks may reflect distinct sampling strategies or prevention policies between the two countries.