<p><strong>Background and objective:</strong> Long‐term respiratory risks following exposure to relatively short periods of poor air quality early in life are unknown. We aimed to evaluate the association between exposure to a 6‐week episode of air pollution from a coal mine fire in children aged <2 years, and their lung function 3 years after the fire.</p> <p><strong>Methods:</strong> We conducted a prospective cohort study. Individual exposure to 24‐h average and peak concentrations of particulate matter with an aerodynamic diameter < 2.5 μm in diameter (PM<sub>2.5</sub>) during the fire were estimated using dispersion and chemical transport modelling. Lung function was measured using the forced oscillation technique (FOT), generating standardized Z‐scores for resistance and reactance at a frequency of 5 Hz (Rrs<sub>5</sub> and Xrs<sub>5</sub>), and area under the reactance curve (AX). We used linear regression models to assess the associations between PM<sub>2.5</sub> exposure and lung function, adjusted for potential confounders.</p> <p><strong>Results:</strong> Of the 203 infants originally recruited, 84 aged 4.3 ± 0.5 years completed FOT testing. Median (interquartile range, IQR) for average and peak PM<sub>2.5</sub> were 7.9 (6.8-16.8) and 103.4 (60.6-150.7) μg/m<sup>3</sup>, respectively. The mean ± SD Z‐scores for Rrs<sub>5</sub>, Xrs<sub>5</sub> and AX were 0.56 ± 0.80, -0.76 ± 0.88 and 0.72 ± 0.92, respectively. After adjustment for potential confounders including maternal smoking during pregnancy, a 10 μg/m<sup>3</sup> increase in average PM<sub>2.5</sub> was significantly associated with worsening AX (β‐coefficient: 0.260; 95% CI: 0.019, 0.502), while the association between a 100‐μg/m<sup>3</sup> increase in peak PM<sub>2.5</sub> and AX was borderline (0.166; 95% CI: -0.002, 0.334).</p> <p><strong>Conclusion:</strong> Infant exposure to coal mine fire emissions could be associated with long‐term impairment of lung reactance.</p>