The purpose of this study was to demonstrate the application of the many-faceted Rasch model to a personality measure. The authors use the model to calibrate the Self-Talk Scale (STS). Good model–data fit supported the measurement of self-talk frequency in adults as a unidimensional construct. Results also supported the proper functioning of the original five-category STS response format. Because of evidence that different items do not contribute equally to the total score, the authors provide information for converting raw STS total scores into more appropriate logit scores. The methodology and results demonstrate how the Rasch model can provide additional support for the validity of measures. Implications for using the Rasch model for personality assessment in general and for using the STS in particular are discussed.