The correct recognition of spread patterns of the COVID-19 virus is crucial for effective decision-making on preventive measures to be taken and protection of society. A forecast of hazard development, substantiated by reliable virus propagation models, can support control of the infection risk and minimize its impact on the population. Decisions to close or limit social activities or to lift bans can then be made in the right area at the right time. This study proposes an iterative agent-based and compartmental model used to analyse the COVID‑19 virus spread. The contagion mechanism is based on population density data and Monte Carlo methods. The model parameters were optimized and verified using data about incidence in Poland between March 3 and November 24, 2020. To obtain optimum values of model parameters, a simulation and analysis were completed for three vaccination strategies begun on December 27, 2020. This study contains numerical results and spatial analyses that represent and confirm the accuracy of assumptions of the proposed model. Certain limitations of the proposed approach and planned directions of future work are also indicated in the final section.