Past research has found that very low scorers on the Beck Depression Inventory (BDI) tend to respond in a "fake-good" manner on the Minnesota Multiphasic Personality Inventory (MMPI) validity scales. This findings was interpreted as evidence of poor "low-end specificity." This study replicated and extended this work by evaluating the low-end specificity of four popular depression measures. The evidence of problems with low-end specificity was strong for the BDI and marginal for the Profile of Mood States depression subscale. Interestingly, MMPI scores in the "normal" range were associated with "fake-good" responding, whereas very low and high scores were not. There was no evidence of an association between fake-good responding and scores on Exner's Depression Index for the Rorschach. Implications of these findings and guidelines for dealing with problems with low-end specificity in research are discussed.