This paper presents an extensive, quantitative comparison of four multiobjective evolutionary algorithms (EAs) applied to the 0/1 knapsack problem, a NP-hard problem that represents an important class of real-world problems. The study aims to evaluate the effectiveness of EAs in finding multiple Pareto-optimal solutions, while also considering the distribution of these solutions. The comparison uses two complementary quantitative measures and includes four different multiobjective EAs as well as a pure random search algorithm. The results show that EAs are particularly suited for multiobjective optimization, as they can process a set of solutions in parallel and exploit similarities between solutions through crossover. However, the study highlights that the performance of EAs is influenced not only by their ability to find solutions but also by the distribution of these solutions. The paper also discusses the different types of multiobjective EAs, including aggregation methods, population-based non-Pareto approaches, and Pareto-based approaches. It emphasizes the importance of preserving diversity in multiobjective optimization and the use of niching techniques, such as fitness sharing, to achieve this. The study concludes that while EAs are effective in finding multiple Pareto-optimal solutions, further research is needed to improve their performance and adaptability to different problem types. The paper is organized into sections that provide an overview of multiobjective EAs, describe the 0/1 knapsack problem and the test data sets used, present the experimental results, and discuss the conclusions and future perspectives of the study.This paper presents an extensive, quantitative comparison of four multiobjective evolutionary algorithms (EAs) applied to the 0/1 knapsack problem, a NP-hard problem that represents an important class of real-world problems. The study aims to evaluate the effectiveness of EAs in finding multiple Pareto-optimal solutions, while also considering the distribution of these solutions. The comparison uses two complementary quantitative measures and includes four different multiobjective EAs as well as a pure random search algorithm. The results show that EAs are particularly suited for multiobjective optimization, as they can process a set of solutions in parallel and exploit similarities between solutions through crossover. However, the study highlights that the performance of EAs is influenced not only by their ability to find solutions but also by the distribution of these solutions. The paper also discusses the different types of multiobjective EAs, including aggregation methods, population-based non-Pareto approaches, and Pareto-based approaches. It emphasizes the importance of preserving diversity in multiobjective optimization and the use of niching techniques, such as fitness sharing, to achieve this. The study concludes that while EAs are effective in finding multiple Pareto-optimal solutions, further research is needed to improve their performance and adaptability to different problem types. The paper is organized into sections that provide an overview of multiobjective EAs, describe the 0/1 knapsack problem and the test data sets used, present the experimental results, and discuss the conclusions and future perspectives of the study.