The crucial importance of valid and reliable scores in performance assessment has led to a range of measures taken to reduce rater variability. These include training and monitoring of raters, statistical adjustment of scores, and approaches that advocate the necessity of plural opinions on a performance sample and involve communal rating sessions to resolve rater differences. Rater reliability and scoring validity are not only major concerns of large‐scale language testing, but are equally important in classroom language assessment. Accordingly, this entry discusses ways in which teachers can improve the reliability and validity of the scores they award to their students' speaking and writing test responses.