Diabetes Mellitus is a metabolic disorder which is characterized by chronical high blood glucose. Automatics detection of diabetes mellitus is still challenging. This study detecting diabetes mellitus by using Kernel k-Means algorithm. Kernel k-Means is algorithm which was develop from k-Means algorithm. Kernel k-means use kernel learning that able to handling non linear separable data, it differences with basically k-Means. The performance of Kernel k-Means in detecting diabetes mellitus also is compared with SOM algorithms. The experiment result show that Kernel k-Means has good performance, better than SOM.