AbstractThe problem considered is that of transporting passengers from the ground floor to higher floors in a building with a single lift during the morning up-peak when stopping at every desired destination floor would produce an inadequate service. If the number of stops is not to exceed some specified maximum, then dynamic programming algorithms are given which determine the optimum floors to stop at, either for predetermined fixed stops, or for stops that depend on the destination floors of the passengers in the lift on a particular journey. The algorithms are evaluated and compared with others using simulation techniques and shown to be robust under various demand conditions. The potential application to real but inadequate lift systems is briefly discussed.