Although in 1973 the average nationwide occupancy rate for hotels and motels was 61%, this rate fluctuates widely based on the time of year and the specific hotel or motel. The purpose of this paper is to present a short range forecasting procedure for occupancy rates which (1) is simple to use; (2) does not require significant computer time or storage; (3) adapts to changes, and (4) is easy to maintain and update. The model developed is an extension of Winter’s three factor model to handle two different kinds of seasonality. The proposed model is tested using both hypothetical and actual data. A discussion of how the model output could be used in setting room rates for convention and tourist groups is also presented.