Performance ratings have been shown to be susceptible to various biases. Two of the most influential biases are assimilation and contrast effects. They denote that ratings are biased either toward (assimilation) or away from (contrast) prior reference assessments. Rigorous rater trainings have proven useful to reduce the latter bias, which, however, are rather time‐consuming. A more parsimonious approach achieving a similar purpose may be to optimize the rating material in order to reduce the tendency of being affected by non‐performance‐related information. Behaviorally anchored ratings scales (BARS) appear to be a theoretically sound alternative in contexts in which rater trainings are uncommon or assessment situations occur only rarely. In two experiments, we evaluated whether BARS are able to reduce assimilation and contrast effects using systematically manipulated standardized videotaped selection interviews. While assimilation effects were reduced by 40%, contrast effects were not significantly but descriptively reduced.