Evolutionary multitask optimization (EMTO) has emerged as a novel paradigm in the field of intelligent optimization, significantly enhancing the efficiency of algorithms and the quality of solutions through cross-task knowledge transfer mechanisms. This paper provides a systematic review of the research progress in the EMTO over the past decade, covering three main dimensions: Technical advances, problem classification, and applications. Firstly, the core techniques of evolutionary multitask algorithms are thoroughly analyzed, including the design of evolutionary frameworks, knowledge transfer mechanisms, and the innovation of adaptive evolutionary operators. Secondly, a classification system for multitask optimization problems is established. The key characteristics and solution strategies for typical scenarios such as single-objective, constrained, competitive, multi-objective, and many-task optimization are elaborated in detail. In addition, this paper outlines the functional features of mainstream EMTO-related tool platforms and introduces successful application cases in fields such as path planning, mathematics, machine learning, and computer vision. Finally, the existing challenges in this field are discussed, and future research directions are forecasted to provide technical references and guidance for relevant scholars.