The monograph is devoted to frequency-domain methods of investigation of multivariable [also called multi-input multi-output (MIMO)] control systems in the presence of stochastic signals. Experts engaged in control engineering are interested in statistical methods because real control systems are often subjected to random disturbances. Many important statistical methods of classical feedback control that are based on frequency-domain concepts have not been adequately extended to multivariable case. This especially concerns the methods of statistical linearization, as well as joint statistical and harmonical linearization, which allow one to investigate the dynamics of control systems containing essential nonlinearities. The goal of this book is to provide a unified statistical control theory of linear and nonlinear (stable or limit cycling) MIMO systems as a direct extension of the conventional statistical theory of classical control. The monograph is intended for practitioners and researchers working in industry. It can also be helpful for lecturers on classical and modern control, as well as for graduate and undergraduate students studying the basics of multivariable control.