Prediction is one of the most important techniques in determining the resulting wind speed. The decision to predict is very important, because with predictions can calculate the average wind speed and a good prediction is an accurate prediction. The purpose of this research is to predict the wind speed using the neural network model to determine the wind speed in the following year. To predict wind speed, the author builds a wind speed prediction model using an artificial neural network (ANN) with a backpropagation learning algorithm. The scope of research data collection from BMKG Sleman. Therefore we need a method that is better than the load coefficient method. The results of this study show that the predictions of ANN wind speed in 2019 from January to December are close to the default data from the BMKG. Sleman and stay stable. And the average error percentage of ANN is 5%.