This study aims to develop a system to classify diseases that attack corn leaves. This study used four types of disease, namely: leaf blight (Helminthosporium turcicum), leaf spot (Bipolaris maydis syn), leaf rust (Puccinia polysora) and downy mildew (Peronosclerospora maydis). This study uses 52 data in the form of images. Every image is changed into vector data using Gabor wavelet filter. This study uses the K-Means Clustering method for disease grouping. The data in this study are vector data. This research process goes through the stages of preprocessing, clustering, and accuracy testing. Preprocessing includes Gabor wavelet filters to extract vector data from the original image. Clustering uses K-Means by determining the starting point manually and calculating similarity using Euclidean Distance. Independent testing of accuracy by comparing the system and manual. The highest accuracy is 98% of the 51 correct data using 52 data with 4 data cluster labels.