New metaheuristic optimization methodologies have been proposed in recent years. These methods have proven to be poor at finding multiple solutions despite being effective at locating a global optimum. The most significant difference between multimodal optimization methods is that they find multiple solutions in a time-varying objective function. For some specific applications, the best theoretical solution may not be the best or most feasible due to its restrictions, so it is very important to find multiple solutions, both local and global. Dynamic multimodal optimization based on evolutionary metaphors has received little attention in the literature despite its importance in the field. Meanwhile, using a non-parametric iterative process known as mean shift, a set of samples of a function is used to detect local maxima. In dynamic environments, the location of local maxima is achieved thanks to the most notable characteristics of this process. Considering dynamic optimization problems, this chapter presents the implementation of the mean shift scheme for the detection of local and global optima. In this approach, fitness and density values are considered for possible solutions; this is achieved by modifying the search strategy of mean shift. In addition, information from previous environments is used to improve the convergence process by implementing a competitive memory and a dynamic strategy. The above characteristics of this methodology ensure that most local and global optima are located in dynamic scenarios.