North Sumatra, as one of the largest provinces in Indonesia, has serious challenges related to poverty that require serious attention North Sumatra, as one of the largest provinces in Indonesia, has serious challenges related to poverty that require serious attention and in-depth analysis. Thus, research on poverty levels in this province becomes very relevant and urgent. Therefore, a more in-depth analysis is needed regarding poverty levels in various regions within this province using data mining methods. The data mining approach is a way to gain understanding from large amounts of data. In the context of the problem of poverty levels, data mining has the potential to help reveal patterns that may be hidden in economic and social data. One algorithm that is often applied in clustering analysis is the K-Means algorithm. The K-Means algorithm is a clustering method that is widely used in data analysis and allows grouping data based on similar characteristics, so that it can be used to classify areas with similar levels of poverty. The results of this research show that data mining with the application of the K-Means algorithm can help produce more effective decisions in analyzing clustering of poverty levels in North Sumatra Province involving the use of data over a ten-year period, namely from 2013 to 2022, which records the number of poor people based on District and city. 3 groups were produced, namely 3 levels of poverty, including relatively stable, very vulnerable and vulnerable. Data from 33 districts or cities in North Sumatra resulted in a poverty level clustering of 1 city that was very vulnerable, 4 cities that were vulnerable and 27 cities that were relatively stable.