This study aimed to evaluate the performance of water level classification from CCTV images in agricultural facilities such as reservoirs. Recently, theCCTV system, widely used for facility monitor or disaster detection, can automatically detect and identify people and objects from the images bydeveloping new technologies such as a deep learning system. Accordingly, we applied the ResNet-50 deep learning system based on ConvolutionalNeural Network and analyzed the water level of the agricultural reservoir from CCTV images obtained from TOMS (Total Operation ManagementSystem) of the Korea Rural Community Corporation. As a result, the accuracy of water level detection was improved by excluding night and rainfallCCTV images and applying measures. For example, the error rate significantly decreased from 24.39 % to 1.43 % in the Bakseok reservoir. We believethat the utilization of CCTVs should be further improved when calculating the amount of water supply and establishing a supply plan according tothe integrated water management policy.