In this study, a market segmentation approach is developed and applied to analyze U.S. household cheese purchases. The segmentation is based on household interpurchase time or the hazard rate of purchases. The hazard rate for a household belonging to a given segment is a function of household demographic and marketing-mix variables, and its baseline is assumed to follow a Weibull distribution. The model is flexible and is able to yield increasing, decreasing, or constant hazard rate functions. Four segments have been discovered in the U.S. household cheese purchase market. Two of the segments contain about 40% of the cheese purchase households, which are frequent buyers with an average interpurchase time of 2 weeks. These frequent cheese purchase households are larger in size, with higher income, less proportion of African Americans, and are insensitive to coupons. They are often referenced in the marketing literatures as loyal customers. In contrast, the other two segments contain about 60% of the households, which are infrequent buyers with an average interpurchase time of 6 weeks. These infrequent cheese purchase households are smaller in size, with lower income, higher proportion of African Americans, and are sensitive coupons. The infrequent purchase households are usually the targets of marketing promotions.