North Sumatra Province is a province well known for its rice field production. But from all regencies and cities in North Sumatra, its rice production is uneven and unstable, sometimes its production goes up, sometimes down. Therefore, it is needed a research in the form of prediction for Production of Rice Field Crops, so that local government of North Sumatra can make policy as early as possible, so that production of paddy rice field can continue to rise in order to support the achievement of food self-sufficiency. In this study data that will be predicted sourced from the Central Bureau of Statistics of North Sumatra Province from 2012 until 2016. The algorithm used to make this prediction is the Backpropagation algorithm. This algorithm has the ability to remember and make generalizations of what has been there before. There are 5 architectural models used in this research, among others 3-5-1 which later will produce predictions with 78% accuracy rate, 3-7-1 = 70%, 3-10-1 = 82%, 3-15 -1 = 82% and 3-9-1 = 91%. The best architecture of the 5 models is 3-9-1 with 91% accuracy and error rate of 0.001-0.05. It is expected that the results of this study can contribute to the government in determining agricultural policy in the future.Keywords—Application, Backpropagation, Prediction, Production, RiceDOI : 10.5281/zenodo.1402832