Objective: To assess the predictive value of bronchial hyper-responsiveness (BHR) for the subsequent development of respiratory symptoms, airflow limitation and decline in lung function among aluminium smelter workers. Methods: An inception cohort study of new employees at two Australian aluminium smelters was conducted. Participants completed a modified British Medical Research Council respiratory questionnaire, spirometry and a methacholine bronchial challenge test at baseline and at annual follow-up reviews. BHR was defined as PD20 ≤4000 μg. Poisson and mixed effects models were fitted to respiratory symptoms and lung function (forced expiratory volume in 1 s (FEV1) and FEV1/forced vital capacity (FVC)). Results: Baseline interview and lung function testing were completed by 278 workers, who were followed for a median of 4 years. BHR at baseline, present in 82 workers, was not associated with incident wheeze risk ratio (RR)=1.07 (95% CI 0.74 to 1.55) and cough RR=0.78 (95% CI 0.45, 1.35), but there was some increased risk of chest tightness RR=1.40 (95% CI 0.99, 1.98) after adjustment for age, sex, smoking and atopy. BHR at baseline was associated with lower FEV1 and FVC, although the rate of annual decline in FEV1 or FVC was similar between those with or without BHR. The specificity of BHR was 77% for wheeze, 70% for cough and 77% for chest tightness, but the sensitivity was poor, at 33%, 24% and 39%, respectively. Conclusion: Methacholine challenge testing at entry to employment was not sufficiently predictive of later adverse respiratory outcomes, and notwithstanding the study limitations is unlikely to be a useful pre-employment or preplacement screening test in the aluminium smelting industry.