Recently, district heating and cooling plants have been installed increasingly with the redevelopment of urban areas. In order to assist operators to operate the plants rationally from economic and energy saving viewpoints, it is required to develop a computer-aided system for real-time operation of the plants. Fundamental functions necessary for the system are to predict energy demands precisely and to conduct unit startup/shutdown scheduling (unit commitment) rationally. Many effective methods of predicting energy demands have been proposed previously. However, no effective methods have been proposed for unit commitment. In addition, the relationship between energy demand prediction and unit commitment has not been investigated. The objectives of the present research are to develop a computer-aided system for real-time optimal operation of district heating and cooling plants and to in vestigate the relationship between energy demand prediction and unit commitment. An ARIMA model has been used in predicting energy demands for a period considered from the current time. In addition, a method of determining the value of a parameter for considering the uncertainty in energy demands has been presented. On the other hand, an optimal operational planning method in consideration of unit startup/shutdown cost has been used for unit commitment. By this method, the unit on/off status and load allocation for the period are determined to minimize the sum of energy supply and unit startup/shutdown cost, and to satisfy energy demands predicted. To evaluate the validity and effectiveness of the system developed, numerical calculations have been carried out using energy demands predicted and measured through a year for an existing district heating and cooling plant. The following are the main results obtained: 1) The characteristics of the errors between energy demands measured and predicted, which depend on season, time, and time-difference between measurement and prediction, have been clarified. From these characteristics, the uncertainty in energy demands which is taken into account in unit commitment has been estimated. 2) As representatives of the relationship between energy demand prediction and unit commitment, the relationships between the error in cold water demand prediction and the rate of increase in operational cost or the rate of deficit in cold water supply have been clarified. In addition, the trade-off relationship between the increase (rate) in operational cost and the deficit (rate) in cold water supply has been clarified with the uncertainty in cold water demand as a parameter. 3) It has been turned out that to reduce the annual operational cost, it is necessary to conduct unit commitment appropriately not only in summer but also in mid-season, and to use an appropriate value of the parameter for the uncertainty in cold water demand. 4) It has been shown that the unit commitment based on energy demands predicted can determine the unit startup/shutdown strategy appropriate even for energy demands measured.