Optimal sequencing of chillers and boilers based on short-term energy demand forecasts represents an untapped opportunity to improve the energy efficiency of central heating and cooling plants. This paper presents a case study conducted in a heating and cooling plant with four boilers and five chillers. Time-series modeling methods are examined to forecast hourly heating and cooling loads 24 hours in advance. The results indicate that a hybrid of auto regressive moving average and change-point models can parsimoniously predict the heating and cooling loads. A boiler and chiller sequencing scheme that uses the day-ahead load forecasts is determined by using a nonlinear programming solver. The implementation of this sequencing scheme is estimated to result in a 4%reduction in heating and 25% reduction in cooling energy use.