The coexistence of barter and cash sale offers an intriguing theme for economic anthropologists to explore. Two examples of this pattern are found in southern highland Peru. In the marketplaces, villages, and fields around Lake Titicaca, some vendors sell fish for cash and others barter them for potatoes and grains. In the snow capped cordilleras that ring the lake, herders load their llamas and donkeys with dried meat, wool, and textiles and travel from the high grasslands to agricultural valleys often a number of days away and thousands of meters below to barter their products for maize and other agricultural foodstuffs.The presence of barter in a peasant economy may not seem unusual, but close inspection of these cases raises questions. In the case of exchange of fish, why is barter concentrated among the poorer peasants and in the area adjacent to the lake? Why are native fish bartered more often than introduced species? The herders appear more difficult to understand, since they derive income from the sale of wool and could purchase needed foodstuffs. They complain of the rigors of their trips, whether through muddy insect-infested canyons to the upper reaches of tributaries of the Amazon or across bleak arid wastes to the oasis valleys in the Pacific coastal desert. They could obtain more maize with less difficulty by selling instead of bartering more of their products. Why do they reject this opportunity? An explanation might also be sought for the persistence of barter in the face of the expansion of marketplace systems in the highland Andes in the last several decades. Pryor (1977:158) points out one difficulty of barter by stating that "as trade expands, it becomes increasingly difficult for a person wishing to barter A for Z to find another person wishing to barter Z for A, a problem enshrined in the phrase 'the double coincidence of wants.'" The advantages of money as a medium of exchange are numerous: it is divisible and storable and can be exchanged for a wide variety of goods. Why has barter continued at all? Is it to be explained solely as a relic, an example of cultural lag?Economic anthropology offers several ways to examine cases such as these. This article adopts a double task: to explain the patterning of barter and cash sale in the Andes and to use this patterning to evaluate rival approaches in economic anthropology. The latter endeavor is relatively unusual, since debate within economic anthropology has tended to be carried out at a theoretical rather than an empirical level. The particular case in hand, the exchange of fish for money and foodstuffs in the Lake Titicaca area in Peru, is of interest because it lends itself well to the testing of rival explanations. Before undertaking the comparison, this article reviews briefly some developments in economic anthropology that have made such efforts infrequent. It explains the formation of the competing hypotheses. Then it presents some background information on the Lake Titicaca region, discusses data collection, tests the hypotheses, and offers some evaluations of the different approaches.