George Collier has maintained a research program concerned with the controls of eating for forty-three years. In the first ten years, Collier approached the problem within the operant conditioning framework proposed by Skinner. This consisted of the intrameal analysis of the reflex strength of a food-rewarded operant in short sessions in food-deprived rats. At the end of the first decade, Collier shifted his attack on the problem by studying the ecological control of meal patterns by procurement and consummatory costs in free-feeding rats. This paper analyzes the reasons for this apparently abrupt evolution from operant conditioning to operant ecology and concludes that they accumulated over about ten years and were intellectual and personal, not technical or professional. This paper also summarizes the major achievements of the research program of operant ecology, notes some emerging problems with economic explanations of ecological controls, and argues that Collier's robust and quantitative behavioral results are ripe for physiological analysis.