A backpropagation neural network was designed for short term unemployment rate forecasting. The forecasting was performed on Kaohsiung unemployemtn rate to demonstrate the predictive capability of the network. Extensive studies were performed on the effects of various factors such as learning rate and the number of hidden nodes. The monthly unemployment rate from June 1983 to Feb. 1992 was evaluated by neural network model and alternative methods, such as space-time series analysis, univerariate ARMA model and state space model, the utilization of neural model significantly provide better forecasting of unemployment rate than any other alternative methods. Generally, the prediction precision of the neural network is 40% higher than the prediciton made by the other models.