This research investigates the effectiveness of computerized adaptive rating scales (CARS) in comparison to the relatively more common behavioral anchored ratings scales (BARS) format. The current study sought to extend the body of psychometric research of CARS while investigating its potential for use in the employment interview. Using 43 videotaped interviews and supervisor job performance ratings, and constructing a new task-performance based CARS, it was hypothesized that employment interview ratings produced using the CARS format would yield significantly higher predictive validity coefficients than those produced by the BARS format. Results showed that while interview ratings produced in the CARS format were predictive of supervisor job performance ratings, they were not significantly higher than ratings in the BARS format. Academic and applied implications are discussed.