1. INTRODUCTION When cultivating a new export market or attempting to expand its share of an existing market, an exporter is likely to find that it has been preceded by its reputation. Minus a readily identifiable brand name, quality reputation may solely reflect the country of origin. For example, when consumers are presented with a large amount of complex product information, they may use the country of origin as a heuristic in forming product impressions without considering other product attributes. (1) In this article I show that this heuristic decision-making process can affect a firm's quality decision and can therefore become self-fulfilling. Hence, any bias that consumers have may be well founded; however, it is only correct because firms respond rationally to it. Colombia's garment industry provides an interesting example of a self-fulfilling unfavorable quality reputation in international trade. As described by Morawetz (1981), a single Colombian garment firm took a contract (for 50,000 men's suits) that was beyond its capability. The poor-quality result so tarnished the American importer's name that other high-quality importers become wary of Colombian-sewn garments. With the payoff to high-quality production reduced, Colombian garment firms then concentrated on low-quality markets, and the newly found unfavorable reputation was justified. In the model constructed here, consumers see only a noisy signal of the product's quality at the time of purchase; therefore, their posterior beliefs are based on this noisy signal and the country-of-origin reputations. As will be made clear, higher-quality firms benefit more from an increase in their country's reputation, so that a better reputation generates an increase in the percentage of high-quality firms. In this way the reputation becomes self-fulfilling and ex ante identical countries can be correctly perceived as differing in their percentage of high-quality producers. I next introduce an economy with two sectors. One sector produces more complex goods, where high quality is not easily verified and is more essential to the product's performance. These may be technology-intensive products, or they may consist of the design and marketing of the goods produced in the other sector. In particular, high-quality production costs and the consumer's value for high-quality goods is greater in this sector. I show that this sector becomes more profitable as the country-of-origin reputation increases. Hence, if one country's equilibrium reputation is low and the other country's equilibrium reputation is high, then the pattern of specialization and international trade can be determined solely by "reputational comparative advantage." (2) Finally, I show that reputational comparative advantage can overshadow a small technological comparative disadvantage in determining the pattern of international trade. Given this pattern of trade and specialization, based on reputational comparative advantage, a high-reputation firm may expend resources to establish a multinational enterprise (MNE) in the low-reputation country. For example, many MNE products require the differing specializations of both the high-and low-reputation countries. Although differing in motivation, this result accords well with more traditional models of vertical MNE. (3) However, this impetus for MNE (as well as its more traditional counterparts) does not explain why the MNE would choose foreign direct investment (FDI) over a less costly licensing arrangement. (4) The quality uncertainty in this model, along with the difference in equilibrium reputations, allows us to address this question. In particular, appropriable reputational rents are generated by FDI when the high-reputation parent firm procures and internalizes production with a low-reputation host firm. This FDI "breaks" the low-reputation equilibrium of the host firm and generates reputational rents. …