The purpose of this paper is to provide a review and an assessment of the state-of-the-art in the area of statistical and equivalent linearization in the analysis of nonlinear stochastic systems. Included in the discussion are systems subjected to stationary and nonstationary excitations as well as parametrically excited systems. A variety of linearization criteria are discussed together with accuracy associated with these procedures. This review also includes some applications of equivalent linearization to the study of dynamics of mechanical and structural systems subjected to environmental loads such as wave loads and earthquake excitations. Limitations of current procedures of equivalent linearization are mentioned and suggestions for areas of further research are made.