This review focuses on the known theoretical and experimental results in the field of obtaining metal matrix composite materials by processing the melts using physical methods in the conditions of casting and metallurgical processes. The possibilities, advantages and disadvantages of various physical impact methods are considered from the standpoint of their effect on the structural and morphological characteristics, physicomechanical and operational properties of cast composite materials based on aluminum and its alloys. The paper provides a classification and a detailed description of physical methods used for melt processing when obtaining metal matrix composites depending on the melt state during processing (melting, pouring and crystallization) and according to the physical principle of the effects applied (thermal, electromagnetic, cavitation, mechanical, etc). The paper describes a contemporary view of the laws and mechanisms of the effect exerted by melt processing using physical methods on the structure and phase formation processes of as-cast metal matrix composites. The currently known effects of the impact on their structure are described from a qualitative and quantitative point of view, in particular, effects associated with a change in the wettability of particles, their distribution, dispersion and morphology, as well as with a change in the structural state of the matrix material. The paper systematizes the data on the properties of metal matrix composites obtained using physical impacts on the melt during melting and crystallization. The research shows the prospects for the development and practical application of physical impact methods for melts in the production of metal matrix composites based on various matrix materials and reinforcement systems including endogenously, exogenously and integrally reinforced composite materials. Priority areas of theoretical research and experimental development are discussed highlighting discussion areas and issues in the field of obtaining metal matrix composites using physical impacts on melts during melting and crystallization. Areas for future research in this field are proposed based on the systematic analysis of key problems limiting the widespread industrial use of physical methods for melt processing.