International guidelines for treatment outcome analysis of tuberculosis cases have been published and are widely used. They do not, however, fully address the incorporation of multidrug-resistant tuberculosis (MDR-TB) cases. Here we present an approach to cohort analysis of treatment outcomes for all registered TB cases, including MDR-TB cases. We analyzed all new pulmonary smear- and/or culture-positive cases registered in Latvia during 2002. Analysis of treatment outcomes at 24 months after initial case registration showed overall treatment success at 84%. This approach to outcome analysis is possible only for settings where MDR-TB treatment is established.