<![CDATA[ The purpose of this research is to analyze the effect of the original regional income (PAD) on Capital Expenditure. The data used in this study are secondary data sourced from the 2014-2018 APBD Realization Report Documents issued by the Directorate General of Fiscal Balance. The analytical method used is descriptive statistics, K-Means, multiple linear regression. The results of this study are that there are 4 clusters that are formed, namely Cluster 1, there are 2 districts / cities with other potential that have high legitimate PAD, Cluster 2 there is 1 district / city with high potential of Separate Regional Wealth Management Results, Cluster 3 exists 6 Districts / cities with conditions that have very low PAD potential and Cluster 4 have 1 Regency / city with rich Regional Tax and Regional Retribution. The results of the next study were obtained by a constant value of 520031.075, determination coefficient (R2) of 0.188 or 18% with the results of the t-test showing that for X1 the t-value < t-table (0.52422 < 2.412) with sig. (0.60270 > 0.05), X2 obtained the value of t-count < t-table (-1.50343 < 2.412) with sig. (0.13971 > 0.05), X3 obtained a t-count value < t-table (0.38560 < 2.412) with sig. (0.70161 > 0.05), X4 obtained t-count value > t-table (2,97733 > 2,412) with sig. (0.00467 < 0.05) and the results of the F-test obtained by the value of f-count > f-table (3.83542 > 2.61) with the value of sig. (0,00914 < 0,05) means that simultaneously PAD has an effect on Capital Expenditures. Keywords: K-Means, PAD, Regional Wealth, Capital Expenditure]]>
