This is a guidebook to current multicriterion decision methods and their mathematical background. The introduction discusses the basic problem, choice of an action which is to do well by several criteria, and the classification of methods. The standard classification is into methods which first calculate a single multiattribute utility function and then maximize that, methods which build a binary relation starting with the strongest preferences of the decisionmaker, and interactive methods, in which the decisionmaker enters additional information about his preferences at successive stages. Chapter 1 discusses examples in relation to the listing of the set of actions. Chapter 2 discusses representation of preferences, for instance by ‘pseudoorders’ which have strong preference, weak preference, and indifference. Chapter 3 takes up concepts such as efficiency and dominance. Chapter 4 takes up various methods of forming a multiattribute utility function. This often involves the decisionmaker considering choices between various combinations of actions. There is also the analytic hierarchy process which involves matrix eigenvalues. Chapter 5 discusses 10 ‘outranking methods’ which form preference relations. For instance the tenth, due to d’Avignon and the author, involves giving a probability distribution of values of the criteria for each action, a preference index on those values, and a summation over the criteria, as well as other mathematical calculations. Chapter 6 goes into interactive methods. For instance, if a decisionmaker is not satisfied at some point, he can make a concession on some criterion. Chapter 7 goes into methods of the nature of preference aggregation. I recommend this book to those involved in the field. F. W. Roush Mathematical Social Sciences Alabama State University Montgomery, AL 36101, USA