an econometric model was built for predictive purposes and for understanding the relationship between the us market and canadian newsprint producers. in this study, a simultaneous-equation model was developed, which consists of four equations, one each for the supply, export, consumption and price of newsprint. the period of 1955-1986 was covered by this model. the data for the years 1985 and 1986 were reserved to test the predictive power of the model. in fitting the four equations for the period of 1955-1984, the coefficients of determinations, the r-square values, between observed and predicted values were higher than 99 per cent. the test results of the forecasting power showed that there was no difference between predicted and observed values at the 5 per cent level of significance. sources of forecasting error are expressed as three partial inequality coefficients associated with bias, variance and covariance of predicted and observed values. in the present forecast, the major source of error came from the covariance between the predicted and observed values. this could not be further reduced. the model can be used to make annual predictions and provides an accurate means of predicting demand, supply, consumption and price of canadian newsprint exports to the united states of america.