Background: Patients with pulmonary tuberculosis are likely to develop pulmonary impairment after tuberculosis (PIAT). The stability of PIAT and the relationship of PIAT to the duration of delay in tuberculosis diagnosis and treatment have not been fully characterized. Methods: We performed serial pulmonary function tests (PFTs) in a cohort treated for pulmonary tuberculosis after 20 weeks of tuberculosis therapy and again on or after treatment completion to determine the stability of PIAT. PFTs were compared with the duration of delay in tuberculosis diagnosis and treatment, as well as other demographic variables. Results: The median duration between the first and second tests was 15 (interquartile range 9-34) weeks. The mean change in FVC was -0.02. l (95% confidence interval [CI] -0.09, 0.06), and the % predicted was -0.02 (95% CI -2.17, 2.12). FEV1 changes were 0. l (95% CI -0.05, 0.06), and the % predicted was -0.11 (95% CI -1.82, 1.60). PIAT was not related to the duration of delay in tuberculosis diagnosis or treatment, age or smoking. Conclusions: PIAT was not associated with the duration of delay in tuberculosis diagnosis and treatment and did not significantly change during follow-up. These data demonstrate that, for many individuals, the completion of tuberculosis treatment is the beginning, not the end, of their tuberculosis illness. © 2011 .