We conducted two studies that assessed the role of initial anxiety in rate of change (depression reduction) in cognitive therapy for major depression. In both studies, depression and anxiety were assessed at intake, and depression was assessed at every treatment session. Longitudinal growth modeling was used to predict rate of change in treatment from sessions 1–12 controlling for intake depression, with intake anxiety as a moderator of change. In Study 1, high initial anxiety was associated with a faster rate of depression reduction across the course of cognitive therapy, whereas in Study 2, high initial anxiety was associated with a faster rate of depression reduction in the early sessions of treatment. The influence of intake depression on rate of change was controlled, and therefore the results are likely not due to greater symptom severity or distress among those high in anxiety. BAI subscale analyses suggest that the results are likely due to the physiological arousal characteristic of anxiety. These results suggest a potentially beneficial role for initial anxiety in cognitive therapy for depression.