We evaluated two methods for characterizing clinically significant change in agoraphobia treatment research: The method proposed by Jacobson and colleagues (1991) and an alternative method, in which the endstate functioning of patients (low, medium, or high) was assessed by several criteria relevant for panic disorder with agoraphobia. Whenever possible, statistically determined cutoff points were applied on these criteria. Comparison of the outcome revealed considerable consonance between both methods, although the Jacobson approach was somewhat more lenient in considering patients recovered. The reliable change index, an additional criterion proposed by Jacobson et al. in order to assess whether patients had experienced true change as a result of treatment, had little informational value: All patients who met the criterion of clinically meaningful change had reliably changed as well. Moreover, the reliable change index did not discriminate between patients with medium and low clinical endstatus.