Accurate prediction of the macroscopic traffic stream variables is essential for traffic operation and management in an intelligent transportation system. Adverse weather conditions like fog, rainfall, and snowfall affect the driver’s visibility, vehicle mobility, and road capacity. The rainfall effect on traffic is not directly proportional to the distance between the weather station and the road because of terrain feature constraints. The prolonged rainfall weakens the drainage system, affects soil absorption capability, which causes waterlogging. The weather event has a spatiotemporal correlation with traffic stream variables, as waterlogging on the road due to rainfall affects the traffic on adjacent roads. The spatiotemporal and prolonged impact of rainfall is not studied in the literature. In this research, we examine whether the inclusion of the rainfall variable improves the traffic stream variables prediction of a deep learning model or not. We use the RNN and LSTM models to capture the spatiotemporal correlation between traffic and rainfall data using past and current traffic and weather information. To capture the prolonged impact of rainfall more extended past sequence of rainfall data than traffic data is used in this study. The roads prone to waterlogging are more affected due to rainfall compared to freeways. Thus we examine the effect of rain on traffic stream variables prediction for different types of roads. The test experiments show that the inclusion of weather data improves the prediction accuracy of the model. The LSTM outperforms other models to capture the spatiotemporal relationship between the rainfall and traffic stream variables.