An interactive algorithm to solve multiple-objective non-linear programming (MONLP) problems is proposed. In each iteration of the proposed algorithm, the decision-maker is presented with a solution and a set of direction (trade-off) vectors indicating possible trade-offs. Using the decision-maker's preferred trade-off vector, a new current solution and the corresponding trade-off vectors are found. The proposed algorithm is illustrated with a numerical example of a replacement model. Finally, the method is compared with four other interactive multiple-objective algorithms.