SETTING: Denmark. OBJECTIVE: To evaluate tuberculosis (TB) treatment outcome using treatment indicators recommended by the World Health Organization (WHO) and the International Union Against Tuberculosis and Lung Disease (IUATLD). DESIGN: Retrospective cohort analysis of all TB patients notified in 1992 (n = 350). TB related information was registered and analyzed using hospital case records and laboratory data from Statens Serum Institut, Copenhagen. RESULTS: Among the 350 TB patients, cure rate was 25.7%, treatment completion rate 36.0%, death rate 12.9%, failure rate 0.9%, defaulter rate 3.1% and transfer out rate 2.3%. The remaining 19.1% 'other cases' did not fulfil any of the WHO indicators, but could be included in the treatment completion rate. During the 4-year follow-up period, 3% of all patients relapsed and 15% died. CONCLUSION: The aim set by the WHO of a cure rate of at least 95% was not achieved. The two main reasons were 1) the fact that sputum samples were not taken late during treatment, and 2) the death rate. In the future sputum examination for M. tuberculosis late during treatment will be encouraged to obtain evidence of cure, but only if the patient can produce sputum, as examination of saliva is of no interest. Death from causes other than TB accounted for 8.3%, making it impossible to attain the desired cure rate in Denmark, and possibly also in other industrialized countries with many elderly TB patients. Additional information on risk factors and methodological issues in performing a TB cohort analysis in a low incidence country is presented.