This study uses computerized personnel microdata on the white male managerial and professional employees at a major U.S. corporation to address the following question: Can the additional earnings which are associated with more labor market experience at a point in time really be explained by higher productivity at the same point in time? Our answer to this question, based on both cross-sectional and longitudinal information, is that performance plays a substantially smaller role in explaining cross-sectional experience-earnings differentials and earnings growth than is claimed by those who have adopted the human capital explanation of the experience-earnings profile. This response depends critically on our assumption that the performance ratings which supervisors give to their white male managerial and professional subordinates adequately reflect the subordinates' relative productivity in the year of assessment; we present a great deal of evidence which strongly supports this assumption.