The authors developed a computer-administered form of the Hamilton Anxiety Scale (HAS; Hamilton, 1959), designed to provide a high degree of correspondence with the clinician interview version of the HAS. Both computer and clinician forms of the HAS were administered to psychiatric outpatients (n = 214) and community-based adults (n = 78). The computer-administered HAS demonstrated high internal consistency (rα = .92) and test–retest reliability (rtt = .96). A correlation of r(290) = .92, p < .001, was found between the computer and the clinician versions. The mean score difference between versions was small but significant, F(1, 290) = 21.36, p < .001. In subjects with anxiety disorders the mean score difference between computer and clinician versions was not significant. Results support the reliability and validity of the computer-administered HAS as an alternative to the clinician-administered version of this measure.