BackgroundIn Ethiopia uncomplicated severe acute malnutrition (SAM) is managed at health posts level through the outpatient therapeutic program (OTP). Yet, evidence on the treatment success rate of the program is scarce. This study determines the treatment outcomes and predictors of time-to-recovery among children 6–59 months of age with SAM managed at the health posts level in Shebedino district, Southern Ethiopia.MethodsThis was a prospective cohort study that enrolled 216 children with SAM identified through a campaign conducted in May 2015 and treated over eight weeks at 25 health posts of the district. The average time-to-recovery was estimated using Kaplan-Meier survival curve and the independent predictors of the recovery were determined using multivariable Cox-proportional hazard model. The outputs of the analyses are presented via adjusted hazard ratio with 95% confidence intervals (AHR, CI).ResultsAt the end of the eight weeks of treatment 79.6% (95% CI: 74.2–85.0%) of cases recovered from SAM with a weight gain rate of 5.4 g/kg/day. The median time-to-recover was 36 days. The analysis indicated, maternal illiteracy (0.54, 0.38–0.78), severe household food insecurity (0.47, 0.28–0.79), walking for more than 1 h to receive the treatment (0.69, 0.50–0.96), diarrhoea co-morbidity (0.63, 0.42–0.91) and practicing sharing of ready to use therapeutic food (RUTF) (0.53, 0.32–0.88) were associated with slower propensity of recovery from SAM. Children who were enrolled with marasmus diagnosis showed lower recovery than children with kwashiorkor (0.30, 0.18–0.51).ConclusionThe median time-to-recover was 36 days. Discouraging sharing of RUTF, appropriate management of diarrhoea in SAM cases and improving access to OTP sites can help to improve the treatment outcome for SAM.Electronic supplementary materialThe online version of this article (10.1186/s12887-019-1407-9) contains supplementary material, which is available to authorized users.