Abstract The occurrence of occupational injuries differs by sex. One of the main explanations put forward is the existence of clear differences in exposure in the workplace, stemming in particular from differences in the jobs held by men and women and the nature of the tasks involved. However, there is a lack of solid data to support this claim, and most traditional workplace analyses have been limited to reporting risks based on sex, but without identifying the reasons that might explain the observed risk differences. The Quebec government, the World Health Organization and the Canadian Institutes of Health Research all recommend conducting sex-differentiated analyses. Although a few studies have incorporated gender into the analysis of musculoskeletal problems at work, so far only two average-quality studies have compared male and female exposure to chemicals, taking occupation and economic sector into account. The objective of our study was to explore the existence of differences in occupational exposure between men and women using existing epidemiological databases. Two epidemiological studies done in Montreal in the late 1990s provided the exposure data; they investigated the relationship between environmental risk factors (including occupational environment) and cancer (lung cancer and breast cancer). Based on subjects’ work history, experts assigned one or more exposures, from a list of 243 possible substances, to each job occupied by 1,657 men and 2,073 women. In our study, the occupational exposure associated with jobs held by men was compared with that of jobs held by women (all occupations and industries combined) to reveal any exposure differences. Exposures were subsequently compared between the jobs held by both sexes within the same occupational groups, then the same occupations and finally the same “occupational group/industry group” pairs. For the purpose of the comparisons, the agreement between the exposure of men’s jobs and that of women’s jobs was calculated using intraclass correlation coefficients. Then, “notable difference” proportions were calculated from the modelling of the male/female exposure differences by applying hierarchical Bayesian models. As anticipated, given men’s and women’s different employment profiles, the analysis of all occupations and all economic sectors revealed differences in occupational exposure between the jobs held by men and those held by women for a large number of chemicals and some physical agents. Generally speaking, the jobs held by men were found to have higher exposure proportions, especially to motor vehicle exhaust, petroleum fractions, polycyclic aromatic hydrocarbons, building material dust and abrasive dust. In contrast, jobs held by women were found to have greater exposure to fabric and textile fibre dust and to aliphatic aldehydes. Most of these differences in exposure proportions disappeared when the analysis was restricted to a male-female comparison within the same occupational group: out of 4,269 points of comparison for which more than 5% of jobs held by men or women were found to be exposed, only 326 (7.6%) of the points showed notable differences. However, where men’s jobs and women’s jobs involved exposure to a given substance, the time-weighted intensity of exposure was similar. Of the 326 notable differences in exposure proportions between men and women, 187 (57.4%) were due to a lack of precision in the occupational code, 78 (23.9%) to differences in the tasks reported by the subjects and 51 (15.6%) to differences related to the industry in which the work was done. Finally, once occupation and industry were taken into account, only 3.1% of the differences in occupational exposure proportions were left with no obvious explanation. To conclude, sex-differentiated analyses are needed to highlight differences in occupational exposure and injuries, as conducting analyses based on occupation and economic sector alone is not sufficient to reveal the subtle differences in job-associated tasks that are also gender related.