Patient admission scheduling (PAS) is a tasking combinatorial optimization problem where a set of patients is assigned to limited facilities such as rooms, timeslots, and beds subject to satisfying a set of predefined constraints. The investigations into the performance of population-based algorithms that utilized to tackle the PAS problem considered in this paper reveal their weaknesses in obtaining quality solutions that create a space to investigate the performance of another population-based method. Thus, in this paper, an Artificial Bee Colony Algorithm (ABC) is proposed to tackle the formulation of the PAS problem under consideration. It is a class of swarm intelligence metaheuristic algorithms based on the intelligent foraging behaviour of honey bees developed to solve continuous and complex optimization problems. Due to the discretization of the PAS, the continuous nature of the ABC algorithm is changed to cope with the rugged solution space of the PAS. The initial feasible solution to the PAS problem is obtained using the room-oriented approach. Then the ABC algorithm optimizes the feasible solutions with the aid of three neighbourhood structures embedded within the employed bee and the onlooker bee operators of the algorithm. The performance of the proposed ABC algorithm based on three different parameters, the solution number (SN), limit value (LV), and the maximum cycle number (MCN) is evaluated on six standard benchmark datasets of the PAS. Two of these main parameters (i.e. SN and LV) are fine-tuned to obtain the best solutions on instances like Test-data 1 = 679.80, Test-data 2 = 1180.40, Test-data 3 = 787.40, Test-data 4 = 1198.60, Test-data 5 = 636.80, and Test-data 6 = 818.60. The best solutions obtained by the proposed method are evaluated against the results of the 19 comparative algorithms comprising five population-based methods, eleven heuristic, and hyperheuristic-based methods, and three integer programming-based methods. The proposed method shows its supremacy in the performance by achieving the best results in all the instances of the dataset when compared with five population-based methods (DFPA, HSA, MBBO-GBS, BBO-GBS, and BBO-RBS) and producing the best results in five instances when compared with eleven heuristic and hyperheuristic-based methods (LAHC, DHS-GD, HTS, DHS-SA, ADAPTIVE GD, GD, HH-GD, DHS-IO, HH-SA, HH-IE, TA) and Finally, it had a competitive performance with the other three Integer programming methods (MIP warm start, MIP-Heuristic, CG) that worked on the same formulations of the PAS. In a nutshell, the proposed ABC algorithm could be adopted as a new template algorithm for the PAS community.