The emergence of an infectious disease pandemic may result in the introduction of restrictions in the distance and number of employees, as was the case of COVID-19 in 2020/2021. In the face of fluctuating restrictions, the process of determining seating plans in office space requires repetitive execution of seat assignments, and manual planning becomes a time-consuming and error-prone task. In this paper, we introduce the Epidemiology-constrained Seating Plan problem (ESP), and we show that it, in general, belongs to the NP-complete class. However, due to some regularities in input data that could a affect computational complexity for practical cases, we conduct experiments for generated test cases. For that reason, we developed a computational environment, including the test case generator, and we published generated benchmarking test cases. Our results show that the problem can be solved to optimality by CPLEX solver only for specific settings, even in regular cases. Therefore, there is a need for new algorithms that could optimize seating plans in more general cases.