Middle-Income Trap (MIT) is a growing concern for developing countries' governments and development banks. This paper uses a mixed approach to identify episodes of MIT in a panel of 132 countries over 1970-2010 and tests empirically three groups of explaining factors advanced by the recent literature on this issue: growth and accumulation regimes, transformation of trade and productive structures and distributional conflicts. We find that the demographic dividend, the patterns of productive change and diversification, skill misallocation, and conflicts help explain why some middle countries underwent persistent growth slowdowns. On the contrary, such frequently evoked explanations as physical and human capital accumulation, democracy, inequality and redistribution did not help differentiating between the middle-income countries caught in the trap and the others. Cumulative effects are also identified since the weakness of employment and innovation capacities dampens the positive effect of human capital and workforce on medium-run growth.