This study investigated the psychometric properties of a newly developed 20-item instrument that assesses one's anticipation of being able to change anxiety: the Anxiety Change Expectancy Scale (ACES). Study 1 evaluated the ACES in undergraduate university students, self-identified as experiencing difficulties with anxiety. Study 2 examined the ACES in a community sample of persons with anxiety difficulties. Study 3 tested the utility of the ACES in predicting treatment change in a group of individuals with generalized anxiety disorder participating in group cognitive behavioral therapy for anxiety. Across these samples, the ACES demonstrated excellent internal reliability (coefficient alphas=.89-.92) as well as good convergent, divergent, and factorial validity. The ACES was also significantly predictive of treatment-related changes in somatic anxiety symptoms and worry. The results of these studies provide strong support for the ACES as a reliable and valid measure of expectancies for changing anxiety.