Objectives; Few Canadian data sources allow the examination of disparities by ethnicity, language, or immigrant status in occupational exposures or health outcomes. However, it is possible to document the mechanisms that can create disparities, such as the over-representation of population groups in high-risk jobs. We evaluated, in the Montréal context, the relationship between the social composition of jobs and their associated risk level.

Methods; We used data from the 2001 Statistics Canada census and from Québec's workers' compensation board for 2000–2002 to characterize job categories defined as major industrial groups crossed with three professional categories (manual, mixed, non-manual). Immigrant, visible, and linguistic minority status variables were used to describe job composition. The frequency rate of compensated health problems and the average duration of compensation determined job risk level. The relationship between the social composition and risk level of jobs was evaluated with Kendall correlations.

Results; The proportion of immigrants and minorities was positively and significantly linked to the risk level across job categories. Many relationships were significant for women only. In analyses done within manual jobs, relationships with the frequency rate reversed and were significant, except for the relationship with the proportion of individuals with knowledge of French only, which remained positive.

Conclusions; Immigrants, visible, and linguistic minorities in Montréal are more likely to work where there is an increased level of compensated risk. Reversed relationships within manual jobs may be explained by under-reporting and under-compensation in vulnerable populations compared to those with knowledge of the province's majority language.