The importance of interpersonal trust in exchange relationships is well-established, and trust is known to have two important and distinct underlying forms, one cognitive, the other affective. Despite this, trust studies often model only the cognitive task-related form and omit the affective form. This article achieves two main objectives. First it demonstrates that interpersonal trust is best considered a bidimensional construct consisting of both cognitive and affective components, and that omitting either form of trust from empirical studies is a conceptual and specification error. Second, we reveal that both forms explain a large amount of variance in the quality of exchange relationships. Also, studies omitting affective trust ignore the more potent of the two trust dimensions. In addition, models examining only one form of trust strongly inflate the apparent effects of that form, and are diagnostically unsound. Our results have important theoretical and managerial implications regarding the conceptualization and measurement of interpersonal trust, and its role in improving the effectiveness of exchange relationships.