The purpose of this research is to determine the differences of financial performance and market performance before, as at, and after direct election for local government executive. This research picks 28 Surabaya based listed companies as samples based on purposive sampling method with several criteria during local government election. Method of hypothesis testing uses multivariate analysis of variance (manova) and Kruskal Wallis test with three groups of periods, which are before, as at, and after the direct election for local government executive. The results show that there is no significant financial performance of local based company in Surabaya caused by direct local government election measured by current ratio, working capital turnover ratio, debt to asset and return to asset ratio. The market performance test shows that the there is significant difference of company?s abnormal return before, as at, and after the direct election for local government executive. Furthermore, the abnormal return after the direct election for local government executive was significant higher than as at the election year.