A hybrid model based on the autoregressive integrated moving average(ARIMA) model and the artificial neural network(ANN) model is proposed to model and predict interval-value time series. The interval-valued time series are converted to the mid-point and the half-range series, the forecasting of which is accomplished through a hybrid model, respectively. The evaluation of the ARIMA, ANN and hybrid models is based on the estimation of the average behavior of the mean squared error with synthetic and real interval-valued series in the framework of a Monte Carlo experiment. The experimental results show that the hybrid model is an effective way to improve the forecasting accuracy achieved by any one of the models separately.