Human Development Index (HDI) is a composite indicator (combined), which is linked to several variables. This indicator can be beneficial if done properly use the comparison across time and across regions, so that the relative position of a region to another region can be determined as well as the progress and achievement comparison with other areas may also be covered. In general, indicators are useful as an advocacy tool for formulators and decision makers in each region, particularly with regard to public policy is selected and set. Utilization HDI indicators can be used more widely, especially in the context of regional autonomy emphasis on districts / cities, where most of the powers, functions and duties have been transferred to local government autonomy. Thus monitoring the development of self-government performance can be evaluated. East Java is a province with a number of districts / cities most in Java, consists of 29 counties and 9 cities. Respective districts / cities in East Java has the characteristics of the population, condition of the area and can not be equated wisdom for all areas depending on their individual requirements. Can also informed that, based on the 2008 national HDI ranking of East Java Province was ranked 18 out of 33 provinces in Indonesia. When compared to the HDI provinces in Java, East Java Province just better than Banten province who is ranked 23. In order to help resolve issues relating to equitable development in the health sector, education and the economy, it is necessary information about the grouping of districts / cities in East Java Province. In this study, the grouping of districts / cities in East Java is done by principal component analysis and cluster analysis to the non hierarchical / K-Means. With a non-hierarchical grouping the regions that have similar properties to form a single group. Grouping districts / cities in East Java based IPM variables divided into 2 groups: the high-potential areas in group 1 and low potential areas in group 2. Based on the analysis of data obtained in the regions of the incoming group 1 there were 24 districts/cities and 13 districts/cities rest went in groups of 2.