Standardised tuberculosis (TB) treatment through directly observed therapy (DOT) is available in South Africa, but the level of adherence to standardised TB treatment and its impact on treatment outcomes is unknown. To describe adherence to standardised TB treatment and provision of DOT, and analyse its impact on treatment outcome. We utilised data collected for an evaluation of the South African national TB surveillance system. A treatment regimen was considered appropriate if based on national treatment guidelines. Multivariate log-binomial regression was used to evaluate the association between treatment regimens, including DOT provision, and treatment outcome. Of 1 339 TB cases in the parent evaluation, 598 (44.7%) were excluded from analysis owing to missing outcome or treatment information. The majority (697, 94.1%) of the remaining 741 patients received an appropriate TB regimen. Almost all patients (717, 96.8%) received DOT, 443 (59.8%) throughout the treatment course and 274 (37.0%) during the intensive (256, 34.6%) or continuation (18, 2.4%) phase. Independent predictors of poor outcome were partial DOT (adjusted risk ratio (aRR) 3.1, 95% confidence interval (CI) 2.2 - 4.3) and previous treatment default (aRR 2.3, 95% CI 1.1 - 4.8). Patients who received incomplete DOT or had a history of defaulting from TB treatment had an increased risk of poor outcomes.