Based on data of dairy milk cow in Animal Farms of Boyolali  District, only shows the total amount of dairy milk cow in Boyolali  District. So that Animal Farms of Boyolali  District does not know which areas produce dairy milk cows with large numbers or small. Therefore, an algorithm is needed to facilitate the grouping of potentially dairy milk cow based on milk production data (liter), number of female dairy cows (how many), number of owners and year of production. In this research, using the K-Means algorithm is used to the grouping of potential dairy milk cow producing areas. By using K-Means aims in facilitating the classification of an area that has the greatest potential dairy milk cows, medium and small. The result is an illustration that shows the regional grouping based on dairy milk cow yields, which are 13 districts that have a potency of dairy milk cow (cluster1), 28 districts that have medium potency dairy cows producing (cluster2), and 28 districts less potential dairy milk cows (cluster3). For further research could be carried out the excavation process variation data variables that clustering results produced can be maximized.Keywords: K-Means algorithm, clustering, data mining, dairy milk cows