Workers in the aluminium industry are widely exposed to particulate matter (PM) during several stages in the manufacturing process at exposure concentrations several orders of magnitude higher than environmental standards. In this study we investigate the relationship between occupational exposures to fine particulate matter (with an aerodynamic diameter less than 2.5 µm PM2.5) in a cohort of aluminium workers, while adjusting for the time varying confounding by underlying health status, characteristic of the healthy worker survivor effect. To address this time-varying confounding, we applied the parametric g-formula to longitudinal pulmonary function data from 5271 actively employed aluminium workers in the United States followed between 1996 and 2013. We adjusted for time-varying health status using a composite risk score based on health insurance claims. We simulated values for forced expiratory volume in one second (FEV1) between ages 18 and 63, under hypothetical interventions on exposures to PM2.5; we then compared the trajectory of FEV1 decline with age under each intervention to the observed natural course (that was actually observed). Under a hypothetical intervention limit for annual average daily exposures to PM2.5 of 175 μg/m3 (corresponding to the 25th percentile of observed exposures), the average FEV1 at age 63 was 37.1 mL higher (95% confidence interval (CI): 1.8, 72.4) than that actually observed, corresponding to 1.2% of the average predicted FEV1. Our results suggest that occupational PM2.5 exposure in the aluminium industry accelerates FEV1 decline. Regulatory measures may mitigate loss of pulmonary function over time in the industry.