In this issue of the journal, Armstrong and Gibbs (see page 740) report the updated results from a mortality study of a large cohort of Quebec aluminium smelter workers.1 2 The focus of their analysis is the quantitative relationship between exposure to polycyclic aromatic hydrocarbons (PAHs) and mortality from lung cancer. Ten more years of follow-up were added along with new cohorts from additional smaller plants for comparison with previously published data3; the study is now the largest ever conducted in aluminium smelting, with 16 431 individuals followed since as far back as 1950 and 677 observed lung cancer deaths.4 More importantly, individual job histories and smoking habits have been recovered, and historical measurements of exposure to benzo[a]pyrene (BaP) and benzene-soluble matter (BSM) could be used to develop a job-exposure matrix (JEM). By applying the JEM, cumulative exposures to BaP and BSM were calculated at the individual level.5 Thus, both statistical power and high quality exposure data were obtained by Armstrong and Gibbs. A new estimate of the slope (and …