The accuracy of short-term air traffic flow forecast is of great significance for the accurate implementation of air traffic management. In order to improve the accuracy of short-term air traffic flow prediction and make full use of historical operation data, this paper proposes a short-term air traffic flow prediction method based on DTW-LSTM. Firstly, the spatio-temporal correlation of air traffic flow is analyzed, based on this feature the spatial correlation between sectors is measured by DTW algorithm. Then, different sectors’s datas were selected and combined according to the spatial correlation measurement results to construct data sets with different input time series lengths, which were input into the LSTM model for training. Finally, the prediction results of different spatio-temporal parameter combination models are analyzed and compared with the prediction results of SVR model without considering spatio-temporal correlation and LSTM model. The experimental results show that compared with the traditional method, the proposed short-term air traffic flow prediction method improves the accuracy of the prediction results by considering the temporal and spatial correlation of the traffic flow, and has a certain improvement compared with other model methods.