Research on method effects in attribute elicitation suggests that different techniques may elicit different attribute sets, which in turn are differentially associated with self-reported attitudes and purchase intentions. However, past research has failed to distinguish between elicitation of perceptual and evaluative attributes. In this paper we focus on elicitation of evaluative attributes. Since evaluation processes typically involve a set of few, but salient, attributes, substantial elicitation method effects are not expected. The present paper reports the results of an experimental comparison of three different techniques (including two variants of one of the techniques) in which the elicitation context was held constant across techniques (product evaluation context). The performance of the different elicitation techniques is compared on common output-dimensions (e.g., attribute importance and predictive ability) and procedural dimensions (task ambiguity and task congruity). The results indicate that method bias is not a serious problem in the elicitation of evaluative attributes. Theoretical and practical implications are discussed.