The applicability of matching analysis, pioneered in the context of laboratory experiments, to the investigation and interpretation of consumer choice in natural environments is explored by the examination of sequential purchases of four product categories based on information from a panel of British consumers. Over a 52-week period, participants recorded data on the brands, quantities, and prices of their purchases. Matching analysis was employed, and individual and aggregated results showed that consumers generally behave according to the predictions of the matching law when the data is averaged over 3- and 5-week periods. The matching results were assessed in light of the results of a substitutability scale, which allowed consumers to state the perceived level of substitutability, independence, and complementarity for 13 product combinations. The results are generally supportive of the predictions made based on the matching results; however, the antimatching that was expected to characterize purchase patterns for complementary products was not observed. (Contains 6 tables and 6 figures.)