Suppose that the coefficient of objective function, the'coefficient matrix and the right-hand side constraints are not d~tennined exactly, but are only known tolie within some real intervals. Such a decision problem which represents a family oflinear programming problems is called an interval linear programming problem In this paper we propose" ~·new concept of solving the linear programming problems with interval coefficients .. Our proposalis based on the notion of the range of the uniquely determined optimal value of eaCh linear programming problem in the family over the interval data. We drive formulae for computing this range as tlJ.e cornerstone of our approach which will allow the decision maker to have a general view of the problem and to choose the rmal solution in a way that best suits his preferences.