BACKGROUND: India is replacing culture and drug sensitivity testing (CDST) with rapid molecular tests for diagnosing MDR-TB. We assessed the impact of rapid tests on time to initiation of treatment and outcomes in patients with MDR-TB compared with CDST. METHODS: A retrospective cohort study involving MDR-TB patients from six districts in Tamil Nadu state, who underwent CDST (2010-2011) and rapid tests (2012-2013). RESULTS: There were 135 patients in the CDST group and 389 in the rapid diagnostic test group. Median time from sputum receipt at the laboratory to initiation of MDR-TB treatment was 130 days (IQR 75-213) in the CDST group and 22 days (IQR 14-38) in the rapid diagnostic test group (p<0.001). Overall treatment success was 30% with CDST and 41% with rapid tests (p<0.05), but there was high loss to follow-up >30% in both groups and missing data were higher in CDST (13%) compared with rapid tests (3%). There were significantly higher risks of unfavourable treatment outcomes in males (aRR 1.3, 95% CI 1.1-1.5) and those with treatment initiation delays >30 days (aRR 1.3, 95% CI 1.0-1.6). CONCLUSION: Rapid molecular diagnostic tests shortened the time to initiate treatment which was associated with reduced unfavourable outcomes in MDR-TB patients. This supports the policy to scale up these tests in India.