The possibility of forecasting seeing quality from routine meteorological data used as input to a numerical model of the boundary layer is discussed after demonstrating examples of reasonably good agreement between observed and calculated seeing on three nights when observed image diameter did not exceed 2 arc sec. Under poor seeing conditions (5 arc sec) agreement was not achieved. The comparisons were made between optical measurements by the Anglo-Australian Observatory and calculations based on temperature soundings from an aircraft and an empirical relationship for estimating dome seeing effects. It is argued that even limited seeing forecasts would assist in optimizing the observing schedules of highly utilized telescopes. A numerical model appears useful to predict C2N profiles of the atmosphere but cannot handle dome seeing effects, which often constitute a significant contribution to overall image quality.