Summary The occupational health and safety statistical indicators that Quebec’s Institut de recherche Robert-Sauvé en santé et en sécurité du travail (IRSST) has been producing for many years show that the average duration of compensation for workplace injuries that led to time-loss compensation (TLC) has almost doubled in 15 years, going from 57 days in 1995–97 to 101 days in 2010–12. While the difference in average duration of compensation between women and men has narrowed over the years (women having always had a higher level than men), the differences between age groups have changed very little. The purpose of this study was to identify, for the period 2005–12, the factors most strongly related to the differences observed in average duration of compensation, by sex and age group, as well as those associated with the increase in average duration of compensation over time. The analyses are based primarily on comparison tests (t test and ANOVA) for average duration of compensation, as well as on multivariate linear regressions. In addition to sex, age group and year injury occurred, the other factors examined included the type of case (workplace accident or occupational disease), whether there was a musculoskeletal disorder (MSD), whether there was any permanent physical or mental impairment (PPMI), the payment of certain categories of disbursements (a positive amount of disbursement within certain accounting categories, including rehabilitation), the industry, the occupational category and the size of the employer (based on insurable payroll). In 2005–12, Quebec’s Commission des normes, de l’équité, de la santé et de la sécurité du travail (CNESST) acknowledged and accepted over 105,400 occupational injury claims on average per year. Of this number, close to 76% led to TLC, a proportion that ranged from 63% to 80%, depending on sex or age group. Furthermore, men accounted for 69% of injuries with TLC, while 25-to-34-year-olds and 35-to-54-year-olds represented 22% and 50% of cases respectively. Depending on sex or age group, the relative proportion of cases receiving at least 14 days’ compensation varies from 42% to 60%, whereas the proportion for cases having over 365 days ranges between 2% and 12%. Note that the cases with over 365 days account for 60% to 82% of all compensated days. While average duration of compensation is slightly higher for women than for men (97.2 days as opposed to 92.1 days), the variation between age groups is greater. From 45.7 days for 15-to-24-year-olds, average duration of compensation rises to 137.4 days for 55-and-overs. In the analysis aimed at identifying the main factors involved in the observed differences between the sexes and age groups, although there is a statistically significant difference for the vast majority of factors, there is little or no associated effect. The existence of a PPMI, the need for rehabilitation and the accounting category associated with the payment of income replacement benefits (IRBs) or medical assistance fees constitute the factors most strongly associated with the duration of compensation. Multivariate linear regression analysis shows that rehabilitation is the main factor associated with the duration of compensation for occupational injuries with TLC. The second factor is the payment of IRBs owing to the unavailability of a suitable job. Although not considered as a factor as such, the relative distribution (structure) of injuries with TLC specific to each sex or age group, according to one factor or another, sometimes reveals components that explain the observed differences. To this end, through simulations for which the differences in distribution by sex or age group were controlled (eliminated) for the factors analysed, standardized average durations of compensation were calculated. The resulting differences in average duration of compensation between the sexes and age groups are generally smaller than those observed. That indicates that the observed differences in duration between the sexes or age groups are mainly due to a difference in the relative distribution of injuries according to the factors analysed. In the simulation controlling for differences in the proportion of injuries with rehabilitation disbursements, the standardized average duration of compensation for women is slightly lower than that for men, which is the reverse of the observed values. For age groups, although there are still differences, they are smaller than the observed ones. In controlling simultaneously for the structure of several factors, the standardized average duration of compensation for women is about 3 days less than for men, whereas the difference between 15-to-24-year-olds and 55-and-overs is only 19 days, which is far less than the observed 90 days. So the differences in distribution by category, for the factors analysed, explain most of the differences between the compensation durations observed for women and men, and between those for the different age groups. All these simulations show that the greater average duration of compensation for women, compared with men, is strongly associated with a higher proportion of injuries involving rehabilitation disbursements for women (8.4%) than for men (7.3%). This is also true for age groups, with 15-to-24-year-olds having a proportion of injuries with rehabilitation disbursements far below that of 55-and-overs (2.6% versus 11.3%). That being said, it should be noted that the average duration of compensation of cases having had rehabilitation is one of the only categories to have declined from 2005 to 2012, which helped to mitigate the increase in average duration of compensation. As for the analysis of changes in average duration of compensation from 2005 to 2012, the statistical tests show that the strength of the association between the year the injury occurred and the average duration of compensation, while statistically significant, is minimal. Given this result, new standardized average durations were calculated according to the different factors analysed, this time controlling for the change observed between 2005 and 2012 in the relative distribution of injuries by sex or age group. The analysis shows that the change in the proportion of injuries that required rehabilitation disbursements and that in the share of cases with PPMI were the two biggest contributing factors to the increase in average duration of compensation over the years. Overall, if it weren’t for the changes in the distribution of injuries with TLC according to various factors, average duration of compensation by sex or age group would have increased at most by 4 days from 2005 to 2012, rather than by 8 to 20 days as observed. . On January 1, 2016, the Commission de la santé et de la sécurité du travail (CSST), the Commission des normes du travail (CNT) and the Commission de l'équité salariale (CES) were consolidated into one to form the Commission des normes, de l’équité, de la santé et de la sécurité du travail (CNESST).