OBJECTIVES : To compare the hearing of gold miners with and without tuberculosis (TB) to determine the effect of TB and its associated risk profile on hearing. METHODS : Audiological and medical surveillance data of 2698 South African gold miners for 2001-2009 were analysed in a retrospective cohort design. Hearing thresholds for the air conduction frequencies (0.5, 1, 2, 3, 4, 6, 8 kHz) in both ears were analysed together with biographical and occupational data. Subjects were divided into two experimental (single TB treatment, n=911 and multiple TB treatment, n=376) and one control group (n=1411). Comparisons between groups included (1) change from baseline to most recent audiogram, (2) most recent hearing thresholds and (3) most recent thresholds in a subset of noise exposed and unexposed groups. RESULTS : Hearing thresholds for the TB groups were significantly (p<0.01) elevated compared to the control group, after correcting for time between baseline and most recent audiogram, threshold at baseline and age at test. Pair-wise comparisons demonstrated the largest threshold differences between the control and multiple TB group. Changes in mean thresholds across TB treatment groups were independent of noise exposure. Hearing thresholds over time also deteriorated significantly more (p<0.01) in workers with TB (single and multiple treatment) than in workers without TB. CONCLUSION : Gold miners with TB, especially with more than one episode of TB, demonstrate significantly poorer hearing thresholds and more pronounced decline in hearing over time independent of noise exposure. The exact cause is likely a complex interaction between TB, including treatment, and its associated risk profile.