<p>Does socially desirable responding produce exaggerated correlations between self-ratings of personality and positive psychology constructs? We operationalized social desirability as the tendency to react to evaluative content in questionnaire items and related one standard measure and one evaluatively neutralized measure of the Big Five to measures of five key positive psychology constructs (optimism, purpose in life, satisfaction with life, well-being and sense of coherence). Structural equation modelling (<i>N</i> = 439) revealed substantially different relationships between the Big Five and positive psychology constructs when the evaluative factor was accounted for. The results indicated that although the Big Five still predict the positive psychology constructs when the evaluative factor is accounted for, their influence is strongly reduced and the evaluative factor itself a major predictor. Arguably, this suggests that when the evaluative factor is not accounted for, the use of self-ratings invites social desirability to inflate the estimated relationships between personality traits and evaluatively loaded constructs in positive psychology. Alternative interpretations of the evaluative factor are discussed, as well as the importance of probing different solutions to measurement issues surrounding evaluative psychological constructs.</p>