In this paper the method of modeling fuzzy intervals in fuzzy decision-making is presented. Described method makes use of constraint logic programming and it is based on the concept of descriptors. This approach is very general and it is consistent with Zadeh's extension principle and Bellman-Zadeh concept of fuzzy decision making. It fulfills Klir's requisite constraint and deals effectively with a drowning effect too. The idea of descriptors of fuzzy intervals and fuzzy constraints is illustrated with computational example of flexible scheduling problem in which robust for drowning effect schedule is found.