It is easy to overestimate the frequency with which adults actually go to the opera and underestimate the frequency with which they watch TV cartoons on Satur- day mornings, based on their self-reports. (Nunnally & Bernstein, 1994, p. 383) Rynes, Colbert, and Brown (2002) pre- sented the following statement to 959 mem- bers of the Society for Human Resource Management (SHRM): Surveys that di- rectly ask employees how important pay is to them are likely to overestimate pays true im- portance in actual decisions (p. 158). If our interpretation (and that of Rynes et al.) of the research literature is accurate, then the correct true-false answer to the above state- ment is false. In other words, people are more likely to underreport than to overreport the importance of pay as a motivational fac- tor in most situations. Put another way, re- search suggests that pay is much more im- portant in peoples actual choices and behaviors than it is in their self-reports of what motivates them, much like the cartoon viewers mentioned in the quote above. Yet, only 35% of the respondents in the Rynes et al. study answered in a way consistent with research findings (i.e., chose false). Our objective in this article is to show that employee surveys regarding the impor- tance of various factors in motivation gener- ally produce results that are inconsistent with studies of actual employee behavior. In particular, we focus on well-documented findings that employees tend to say that pay