Sometimes tuberculosis diagnosis is missed at hospital admission. Delayed diagnosis of active pulmonary tuberculosis among hospitalized patients could to contribute to nosocomial transmission. The objective of this study was to define the occurrence and associated patient risk factors among hospitalized patients with delayed diagnosis of respiratory tuberculosis. A retrospective chart review was undertaken between 1995 and 2002 on all patients with pulmonary tuberculosis. Time intervals between admission, diagnosis and treatment of tuberculosis were determined. Epidemiological and clinical features were evaluated for their effect on these time intervals. Among 149 patients newly diagnosed to have active pulmonary TB, the diagnosis was initially missed in 102 (68% [95% CI, 61 to 75%]) of all hospitalized patients, of whom 66 (65% [95% CI, 56 to 74%]) were smear positive. Treatment was initiated after a week or more in 62 (42% [95% CI, 34% to 50%]) of all patients, of whom 34 (55% [95% CI, 43 to 67%])) were smear positive. Age >or= 60 years (OR 3.17 [95% CI; 1.27 to 7.87]; p = 0.013), presence of chronic lung disease (OR 2.99 [95% CI; 1.21 to 7.38]; p = 0.017), negative sputum AFB smear (OR 4.51 [95% CI; 1.34 to 15.16]; p = 0.015) and absence of hemoptysis or melanoptysis (OR 2.96 [95% CI; 1.18 a 7.41]; p = 0.020), were independently associated with delays. The diagnosis and treatment of hospitalized patients with pulmonary tuberculosis is often delayed because absence of clinic suspicion owing to old age, chronic lung disease or atypical presentations and slow confirmation by culture. Improved clinical acumen, development of rapid diagnostic tests, and the institution of early empiric therapy are desirable objectives to improve the tuberculosis control.