This study develops a Suspicion Index (SI) to detect the dishonest applicants for direct payment programs more efficiently. Based on the 2014 application data for the three direct payments for rice, upland crops, and less favored areas, we develop the SI through two different approaches: averaging the possibilities of fraud by several farm characteristics and logistic regression. The result shows that the SI is higher for dishonest applicants compared with the others as expected. Therefore, the SI contributes to improving the detection rate of dishonest applicants and their dishonest amounts. The result of this study could contribute to a more efficient administrative process and government budget savings for the direct payment program.