Our thesis explores the drop-out motives of a panel of Peruvian children, provided by the Young Lives program (Boyden, April 2014). We used survival analysis with both, a non-parametric estimation (Kaplan-Meier survival estimates) and a semi-parametric estimation (Cox proportional hazards (PH) model). Our results suggest that the child's initial conditions are relevant determinants of the drop out decision. Also, we found evidence to support the importance of the wealth level, the location of the household and the sex of the caregiver, for the drop-out decision. We believe that further research is required to rene the size of the effects and this can be achieved by the inclusion of the fourth round of data and the use of parametric models to explore the effect of the covariates that did not meet the proportionality assumption required by the Cox PH model.