Gerhart and colleagues (2000) and Huselid and Becker (2000) recently debated the presence and implications of measurement error in measures of human resource practices. This paper presents data from 3 more studies, 1 of large organizations from different industries at the corporate level, 1 from commercial banks, and the other of autonomous business units at the level of the job. Results of all 3 studies provide additional evidence that single respondent measures of HR practices contain large amounts of measurement error. Implications for future research into the HR firm performance relationship are discussed.