Railway track buckling occurs due to inadequate rail stress adjustment. Stress alterations are greatly influenced by the variation in rail temperature due to weather fluctuations. The main aim of this paper is to describe a 24 hour rail temperature prediction model which was developed for use in rail operations to assist in predicting adverse rail temperature conditions which could lead to or initiate track buckling. By applying predictive information, trains can be directed to travel at a regulated speed to reduce the longitudinal loading on the track. The model uses multivariate linear regression to predict the rail temperatures from the predicted weather forecast. Data from a field experiment, involving a weather station and rail temperature sensors, was statistically evaluated and compared to determine the accuracy of the rail temperature model. This paper evaluates the accuracy of i) rail temperatures predicted using only Bureau of Meteorology (BoM) weather forecasts ii) rail temperature predictions using on site weather station data, and iii) empirical weather prediction equations. Concluding results of this paper show that rail temperature can be predicted 24 hour in advance from a BoM forecasts and if calibrated properly the accuracy is within +/- 2.6 ºC of actual rail temperatures. Real time rail temperature predictions using on site weather station data have an accuracy of +/- 4.2 ºC; whereas empirical methods have at most an accuracy of +/- 5.9 ºC from actual rail temperatures. The accuracy of the forecasted rail temperature prediction using BoM weather forecasts is within the magnitude of temperature sensors accuracies this is very encouraging for possibility of rail forecasting rail temperatures without the use of instrumentation.