The entropy generation minimization (EGM) technique is an important tool for the optimization of the thermal systems via the analysis of the associated irreversibilities measured by the entropy generation. This article presents a detailed review of works on the entropy generation analysis during natural convection in various enclosures and processes involving different practical applications. The mathematical formulations of the fundamental governing equations for natural convection followed by the equations of the entropy generation are presented. The calculation procedure of the entropy generation for various test cases is reported briefly with the finite difference and finite volume techniques for some test cases and the detailed discussion of the evaluation of the entropy generation via the finite element method is addressed. Further, the problem formulation and results in terms of the entropy generation are discussed for natural convection in enclosures of various shapes (square/rectangular, trapezoidal, triangular, parallellogrammic/rhombic, curved/wavy). The brief discussion on the entropy generation analysis during various practical applications is also addressed. Overall, the minimum entropy generation vs enhanced heat transfer rate is the main issue in all the case studies with various enclosures involving a number of practical applications to achieve the optimal configuration with the high energy efficiency. The need of the renewable energy is increasing day by day. Thus, the conversion of the renewable energy to a useful form is one of the most challenging processes and natural convection plays an important role in the conversion. The loss of the available energy via the entropy production during natural convection is highly important for the design of suitable energy systems. This review article further provides basis for future research on the entropy generation analysis during natural convection in order to improve the energy efficiency which may be applicable for various renewable energy systems. © 2017 Elsevier Ltd