In order to reduce the effort of road managers, a method for predicting as look-ahead information of road surface conditions based on previously evaluated road surface condition and weather information is proposed. As a first step, by using the previously evaluated road surface condition and weather data, spatial direction series data are constructed. After that, the look-ahead information is predicted based on Long Short Term Memory (LSTM) by utilizing the constructed spatial direction series data. The effectiveness of the proposed method is confirmed by experiment using the actual data which were captured in Hokkaido, Japan.