Theories are useful in predicting events and exploring alternatives only to the extent they accurately describe existing states of affairs. The development of a curriculum theory with predictive and exploratory power depends on accurately describing decision-making systems that result in curriculum -- however curriculum is defined. Three approaches to establishing this empirical base for curriculum theory are discussed: (1) "policy-capturing," a simulation of decision-making processes by means of a regression model; (2) studying decision making as a social influence system; and (3) exploring decision making in terms of information processing or communications systems. The discussion specifies objectives behaviorally, quantifies variables in the decision-making process, and determines the appropriate units of measure for certain qualitative variables. (Page three may reproduce poorly because of marginal legibility). (Author)