The paper has examined the determinants of graduation of MSEs and tried to estimate the average time required to move from one stage into the next stage (i.e., from micro into small and small to medium level). Using 666 nationally representative MSEs sample data from the World Bank Enterprise pool data collected in 2015, both parametric and semi-parametric duration models are applied to estimate the conditional probability of MSEs graduation. The models used include the full parametric models (the Weibull Proportional Hazard (PH model) and the Lognormal Accelerated Failure Time (AFT model) and the Cox Proportional Hazard (Cox PH) model among the semi-parametric models. A number of interesting findings have emerged. Our findings confirm that percent of manager’s time allocated to the enterprise’s affairs, managers’ experience, business/system improvement, enterprise’s location and percent of internal funds/earnings mobilized were found to be positive and significant to increase the likelihood of early graduation of micro or small firms. On the other side, variables, such as degree of competition from informal firms, power supply shortage, land access problem and access to finance from private source increased the graduation duration. It appears that solving power supply shortage and land administration governance is an effective way to increase their graduation. If speedy graduation is desired, then policies should encourage more skill-oriented training about the new innovation and land administration.