This paper presents an analytic (algorithmic) method for determining the optimum proportions of aggregates that satisfy specific gradation, design, and cost requirements of the blend. The gradation requirements include the specification limits and mean deviation from the midpoint of specifications limits. The design requirements include fineness modulus, plasticity index, and specific gravity. The optimum proportions may correspond to minimum mean deviation or minimum cost. In addition to providing the optimum solutions, the method provides the entire feasible region of proportions that may be useful if adjustments of proportions are deemed necessary.The paper first reviews existing aggregate blending methods, classifies them, and describes how the proposed method fits into that classification system. The method is then described along with its graphical interpretation. The method may be regarded as a simulation process and as such is not subjected to any type of constraints. Application of the method to a numerical aggregate blending problem was made to illustrate its use.While existing methods consider minimizing mean deviation (with no consideration of cost) or minimizing cost (with no consideration of mean deviation) the proposed method considers both mean deviation and cost as well as the physical properties. Other advantages of the method include adaptability to linear or stepwise cost functions of aggregates, automatic elimination of negative solutions, and provision for different weights (importance levels) of sieve specifications.Key words: blending, aggregate proportions, cost, mean deviation, gradation, specific gravity, plasticity index, fineness modulus, asphaltic concrete.