Proper maintenance of a scale established in the baseline year would assure the accurate estimation of growth in subsequent years. Scale maintenance is especially important when the state performance standards must be preserved for future administrations. To ensure proper maintenance of a scale, the selection of anchor items and evaluation of their performance in the succeeding administrations are crucial. Under the common item equating design, anchor item selection is a critical process and is typically scrutinized against a set of guidelines for selection. Even if the guidelines for selecting anchor items are met for the anchor set, items may still perform differentially, for example, context effects, or changes in instructional emphasis, which might introduce additional dimensions, hence, differential examinee performance. In this study, four statistical methods of evaluating anchor items are explored. The evaluation process used data from statewide assessments in two content areas. The results indicate that the Delta-Plot method which uses proportion correct (p-value) as input, differs from the methods (Lord's Chi-Square, Stocking and Lord's TCC inverse, and Raju's area minimization methods) that use Item Response Theory (IRT) item parameters as input in flagging differentially performing items. The results further show that the decisions of excluding or including anchor items impacts on students' test scores at both the student and group levels. Keywords: item response theory (IRT), Equating, Anchor. Flagged Items and Their Characteristics is appended. (Contains 5 tables and 2 figures.)