Monitoring the condition of the plant becomes very important. This is caused by the large number of disease attacks and lack of treatment. In addition to the considerable distance for plant treatment, the structure of plant parts affected by the disease needs to be classified. High disease caused by leaves can result in panel failure. So that the delay in the process of diagnosing the disease causes the disease in plants to become severe. This study aims to utilize drone technology and display it in image form to detect disease on plant leaves. We use a neural network to process the results on the image. Neural Network is one of the most effective algorithms in modeling the relationship between input and output data. The results of the study, namely the accuracy of detecting diseases on leaves, taking images from drones reached 99.63