Two models have been developed to derive the optimal room mix for a hotel. The first is a linear programming (L.P.) model, which assumes average utilization factors for the various room types and which finds the room mix for maximum yearly profit under a set of constraints. The second model utilizes an existing model developed by the author for finding the optimal rental policies over time for a hotel having a given room mix, and which provided as a by product the expected contribution to profit per day of activity of the hotel if optimal rental policies wore followed. Thus, by evaluating the expected profit contribution per day for different room type combinations, it is possible to derive a room mix which maximizes the annual rate of return on the investment and which is not distorted by false assumptions as in the ease of the L.P. model. A search procedure has been outlined and it has been suggested that the calculations could be reduced significantly if the search would be limited to a range around the initial L.P. solution.