<p><?tk 2?>In recent years, the production strategy in manufacturing has gradually moved away from make-to-stock toward make-to-order, emphasizing customer orientation. According to production schedules, accepting and processing all potential orders may not be an acceptable decision due to limited resources, which could lead to reduced profitability or even the loss of customers. However, in research studies on flexible job shops, the concept of order acceptance and job rejection policies has often been overlooked. This paper addresses job rejection policy, time-of-use electricity tariffs, dual-resource constraints, and rate-modifying maintenance activities within the context of the multi-objective flexible job-shop scheduling problem. A mixed-integer linear programming model is presented to simultaneously minimize makespan and total earliness. Given the NP-hardness of the problem, this study applies four metaheuristic approaches, including the non-dominated sorting genetic algorithm (NSGA-II), non-dominated ranking genetic algorithm (NRGA), multi-objective invasive weed optimization (MO-IWO), and multi-objective evolutionary algorithms based on decomposition (MOEA/D). The proposed algorithms are developed to solve small, medium, and large-sized instances and their performance is empirically evaluated using these randomly generated test problems. Different performance metrics including the mean ideal distance (MID), the number of non-dominated solutions (NNS), the spread of non-dominated solutions (SNS), and the rate of achievement of objectives simultaneously (RAS) are employed for a comparative analysis of the algorithm results. Computational study indicates that on average across the results, the NSGA-II algorithm outperforms the other algorithms in terms of the MID metric, the MOEA/D algorithm performs better with respect to the NNS and RAS metrics, whereas NRGA demonstrates superiority with respect to the SNS metric.</p>