Disease prediction has a vital role in health informatics. The early detection of diseases assists in taking preventive steps and more functional treatment. Incorporating intelligent classification models and data analysis methods has intrinsic impact on converting such trivial, row data into worthy useful knowledge. Due to the explosion in computational and medical technologies, we observe an explosion in the volume of health- and medical-related data. Medical datasets are high-dimensional datasets, which make the process of building a classification model that searches for optimal set of features a hard, yet more challenging task. Hence, this chapter introduces a fundamental class of optimization known as the multi-objective evolutionary algorithms (MOEA) for optimization, which handles the feature selection for classification in medical applications. The chapter presents an introduction to multi-objective optimization and their related mathematical models. Furthermore, this chapter investigates the utilization of a well-regarded multi-objective particle swarm optimization (MOPSO) as wrapper-based feature selection method, in order to detect the presence or absence of different types of diseases. Therefore, the performance of MOPSO and its behavior are examined by comparing it with other well-regarded MOEAs on several medical datasets. The experimental results on most of the medical datasets show that the MOPSO algorithm outperforms other algorithms such as non-dominated sorting genetic algorithm (NSGA-II) and multi-objective evolutionary algorithm based on decomposition (MOEA/D) in terms of classification accuracy and minimum number of features.