The studies about fragile states in the international relations begins around the 1990s, and the issue remains active. According to the World Bank, in 2016, 1.8 billion people lived in fragile states. The projection shows that, in 2050, almost 10% of the global population will live in fragile contexts. Poverty, famine, child mortality, and violence denote some characteristics of state’ fragility. Many fragile states have been sanctuaries for terrorist groups, threatening global security. There are several indications of causes for fragility in the states, however, there is no universal consensus on this. International organizations, such as the Organization for Economic Co-operation and Development (OECD), find a way to tackle the fragility of states. The problem is: what are the significant variables to justify the defragilization of fragile states, once statistics show that, while one group of states in the world remains fragile or even deteriorated, another group is getting better gradually? The goal of this article is to present the results of research and the outcomes of logit regression analysis, of a group of states, from 2007-2016., which got defragilization. The analysis took into consideration 1,473 relevant observations involving 13 supposed explanatory variables. Some independent variables, such as international aid or investments on education, would appear to be significant, however, they were not. Conversely, the results were consistent with the improvement of state governance and reduction of infant mortality. The study opens an avenue for new researches to defragilization processes.