Objectives There is a paucity of information on environmental risk factors for prostate cancer. We conducted a case–control study in Montreal to estimate associations with exposure to ground-level nitrogen dioxide (NO₂), a marker for traffic-related air pollution.

Methods Cases were 803 men with incident prostate cancer, ≤75 years of age, and diagnosed across all French hospitals in Montreal. Concurrently, 969 controls were drawn from electoral lists of French-speaking individuals residing in the same electoral districts as the cases and frequency-matched by age. Concentrations of NO₂ were measured across Montreal in 2005–2006. We developed a land use regression model to predict concentrations of NO₂ across Montreal for 2006. These estimates were back-extrapolated to 1996. Estimates were linked to residential addresses at the time of diagnosis or interview. Unconditional logistic regression was used, adjusting for potential confounding variables.

Results For each increase of 5 parts per billion of NO₂, as estimated from the original land use regression model in 2006, the OR_5ppb adjusted for personal factors was 1.44 (95% CI 1.21 to 1.73). Adding in contextual factors attenuated the OR_5ppb to 1.27 (95% CI 1.03 to 1.58). One method for back-extrapolating concentrations of NO₂ to 1996 (about 10 years before the index date) gave the following OR_5ppb: 1.41 (95% CI 1.24 to 1.62) when personal factors were included, and 1.30 (95% CI 1.11 to 1.52) when contextual factors were added.

Conclusions Exposure to ambient concentrations of NO₂ at the current address was associated with an increased risk of prostate cancer. This novel finding requires replication.