Interpersonal attraction in a face-to-face setting is examined in relation to attitude similarity-dissimilarity and accomplice for the S. In contrast to previous research (Byrne & Griffitt, 1966), attraction scores in the range were observed. Ss reciprocated the stranger's evaluation and attraction toward them regardless of the degree of attitude similarity. Both reciprocal attraction and similarity attraction effects are discussed in terms of Byrne and Clore's (1967) competence-reinforcement theory of attraction. The association between attitude similarity and interpersonal attraction is well documented in a number of studies (Byrne, 1971). Byrne has argued that similar attitudes function as positive reinforcements; attraction then increases as a linear function of the proportion of similar attitudes (Byrne & Clore, 1967). Aronson and Worchel (1966) have proposed an alternative account of the similarity attraction effect. Rather than reinforcements, similar attitudes function as cues to the likelihood that the stranger would like the S. The individual has presumably learned to expect approval from persons with similar attitudes. The expectancy of being liked elicits attraction toward the similar stranger. In support of this inferred attraction hypothesis, Aronson and Worchel (1966) found that liking had a Significant effect on attraction regardless of attitude similarity-dissimilarity . Subsequent research (Byrne & Griffitt, 1966) indicated that the absence of a relationship between similarity and attraction in the Aronson et al study was attributable to their use of a restricted range of attitude Similarity-dissimilarity . Byrne and Griffitt (1966) concluded, however, that personal evaluations were much more important influences on attraction than similar attitudes. In both the Aronson and Byrne studies the accomplice not only evaluated the S negatively, but also insulted the S by describing him as shallow, uninteresting, and uniformed. In no condition was a mean attraction score less than 8.0, which indicates liking, observed with the interpersonal judgment scale. Two possible accounts are suggested, (a) that Ss found the situation unbelievable or (b) that dislike ratings do not occur in face-to-face paradigms. Research by Gormly (in press) suggests that the latter hypothesis is inadequate, because this investigator reported mean attraction scores in the range in response to face-to-face disagreements. The absence of scores in the range in the Aronson-Byrne studies is probably attributable to S suspicion.