A retrospective study was conducted to assess TB treatment outcome and thus evaluate DOTS programme in Benishangul Gumuz Region, western Ethiopia. The records of 3658 TB patients registered in the two hospitals between November, 2003 and October, 2012 at DOTS Clinics were analyzed. From the total patients, 2,223 (60.77%) were successfully treated, 315 (8.61%) lost to follow up, 341 (9.32%) died, 4 (0.11%) failed treatment and 775 (21.19%) not evaluated. There is no association in treatment success rate between new and previously treated patients (60.55% vs. 58.33%, X 2 =0.14; P>0.704). When transferred out (not evaluated) is excluded, the success rate was increased approximately by a rate of 0.03 times while the time is increased by one year (β=0.03106; CI=0.0218-0.0403). Smear negative pulmonary TB patients had 1.24 times low treatment success rate (64.08% vs. 58.07%; CI=1.02-1.51; P<0.001) compared to smear positive. Patients from Pawe areas had 1.34 times lower treatment success rate compared to cases from Assosa (64.91% vs. 58.57%; CI=1.16-1.55). Not evaluated and lost to follow up rates were higher in Pawe (24.93% vs. 14.16%; X 2 =38.66; P<0.001 and 10.10% vs. 5.82%; X 2 =16.40; P<0.001 respectively) while death rate was higher in Assosa (14.95% vs. 6.33%; X 2 =59.05; P<0.001). Accordingly, males and elders had the trend to be more likely to experience death. In conclusion, the treatment success rate of TB patients was low which accounts for 60.77%. Low treatment success and high proportion of death (9.32%) and defaulted (8.61%) rates are serious public health concerns that point out low DOTs programme performance in the study areas.