Purchase probability as a function of interpurchase time was examined through comparison of findings from laboratory experiments on reinforcement schedules and from marketing investigations of consumers' interpurchase time. Panel data, based on a sample of 80 consumers who purchased nine supermarket food products during 16 weeks, were used. For each product category, interpurchase time was similar for each shopping occasion and cumulative purchase probability increased as a Gamma function of the time since the last purchase. A comparison of interpurchase times across products and consumers showed that average interpurchase time differed across four subsets of products and across seven groups of consumers, with a significant interaction effect. Interpurchase times tended to be longer after larger purchases, as would be predicted from laboratory results. A correlation between individual interpurchase time and number of products bought on each shopping occasion indicated that consumers who shop more frequently buy larger numbers of products per occasion. These results have several managerial implications and demonstrate the usefulness of a behavior-analytic framework in the interpretation of consumer behavior. (Contains 3 tables and 2 figures.) [This study was financially supported by CAPES (Ministry of Education), CNPq (Ministry of Science and Technology), and FINATEC (Fundacao de Empreendimentos Cientificos e Tecnologicos Brasilia, DF).]