The population growth rate in various countries in the world is increasing, including Indonesia. The population explosion as a result of rapid population growth has a negative impact on the socio-economic life of the community, such as increasing unemployment rates, food shortages, and high poverty rates. Therefore, local governments in each country try to overcome the poverty problem using various policies, including in Surabaya, East Java, Indonesia. This study aims to classify villages in Surabaya using non-hierarchical clustering, such as K-Means, Fuzzy C-Means, Fuzzy Gustafson Kessel, and DBSCAN (Density-Based Spatial Clustering of Applications with Noise), based on poverty indicators. Before analysis, the villages in Surabaya, East Java, Indonesia were classified using non-hierarchical clustering, and the results of cluster analysis were compared from various methods using the value of within clusters sum of squares and average silhouette width. Comparison between village grouping methods results in K-Means being the best method for village grouping in Surabaya, East Java, Indonesia based on the values of the within clusters sum of squares. While based on the average silhouette width value, the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) method is found to be the best method because its value was close to 1 compared to the other methods. Thus, it can be implicated that K-Means and DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is the best method for village grouping in Surabaya, East Java, Indonesia in relation to poverty problems.