In today's competitive business environment, companies need to manage their limited resources. The Multi-Floor Facility Layout Problem (MFLP) is an approach to manage limited space and budgets. The goal of MFLP is to determine the placement of facilities in a multi-floor building without any overlapping with the aim of minimizing costs. In this study, a Multi-floor Multi-row Facility Layout Problem (MFMRFLP) model is proposed. The proposed model presented an MFLP with a multi-row layout on each floor. Besides the layout of the facilities, the model also determines the elevator location based on both horizontal and vertical movements. Since the problem is NP-hard, a genetic algorithm (GA) was also employed to solve the problem. The proposed GA is compared against an exact method to evaluate their performances. The results demonstrate the GA's efficiency in solving the MFMRFLP within a reasonable timeframe, outperforming the exact method, particularly in large-scale instances. Specifically, the GA achieved optimal or near-optimal solutions, showing its superior performance in solving complex, real-world facility layout optimization problems.