This research addresses the epidemiological context of the COVID-19 pandemic, emphasizing the specific panorama of Brazil, with a special focus on the State of Parana, approximately one year after the implementation of the first lockdown. The general objective of this research is to apply the clustering technique known as k-means to categorize municipalities in Parana based on two main variables: the daily number of confirmed cases and the number of deaths from COVID-19. To achieve this purpose, data provided by the Parana Department of Health were used, covering the period from January 1, 2021 to March 15, 2021. The results obtained revealed the identification of three clusters that stood out as excellent , highlighting divergences in the incidence patterns of cases and deaths between municipalities. Notably, a correlation was observed between population density and the frequency of cases and deaths, with more densely populated areas tending to record higher numbers. Furthermore, the evaluation of the accuracy of the J48 and Naive Bayes algorithms in classifying clusters demonstrated satisfactory results. Consequently, it is concluded that the grouping technique used proved to be effective in identifying similarities in the spread patterns of COVID-19, offering relevant evidence for the formulation of targeted and efficient strategies to combat the pandemic, especially in the most impacted regions.