CiteScore

CiteScore is a journal level metric calculated by Scopus that measures the average number of citations received per document published in a journal over a four year citation window. Unlike the two year journal impact factor calculated from Web of Science data, CiteScore's longer citation window captures citations that accumulate more slowly, particularly in disciplines where research takes longer to be cited.

CiteScore provides researchers with a standardized way to compare the citation impact of journals within a discipline, and its percentiles support comparisons across disciplines. When selecting a target journal for manuscript submission, researchers can use CiteScore alongside other factors (scope, audience, turnaround time, acceptance rate) to gauge a journal's visibility and impact. For librarians and institutions, CiteScore data inform collection development and subscription decisions. CiteScore is freely available on the Scopus website, making it more accessible than the journal impact factor, which requires a subscription to Journal Citation Reports.

Since Elsevier changed the method in 2020, CiteScore has used the same four year window for citations and for documents. It counts the citations received in the calculation year and the three years before it by documents the journal published in those same four years, and divides that count by the number of those documents. For example, the 2024 CiteScore of a journal counts the citations received in 2021, 2022, 2023, and 2024 by articles, reviews, conference papers, book chapters, and data papers published in 2021, 2022, 2023, and 2024, divided by the number of those documents published in the same four years.

The four year window is a deliberate design choice that distinguishes CiteScore from the journal impact factor, which uses a two year window. The longer window benefits journals in fields like mathematics, engineering, and the social sciences, where citations accumulate more slowly. It also reduces the volatility of the metric, since individual highly cited papers have less influence on the calculation when spread over four years of publication.

Only these five peer reviewed document types count, and the same types are used in both the numerator and the denominator; editorials, letters, and news items are excluded from both. Before 2020, CiteScore used a three year window and counted all document types, including letters and editorials. The matched approach contrasts with the journal impact factor, which includes citations to all document types in the numerator but only "citable items" (articles and reviews) in the denominator, a mismatch that has been criticized for inconsistency and for potentially inflating impact factor values.

Scopus also publishes the CiteScore Tracker, which provides a running monthly estimate of the upcoming CiteScore, allowing users to see how a journal's performance is trending during the current year. CiteScore percentiles rank journals within their subject category, facilitating comparisons across disciplines.

CiteScore is useful when comparing journal impact within a discipline, when evaluating journals in fields where citation accumulation is slow, when a freely available metric is preferred, and when a metric that counts the same document types in its numerator and denominator is desired. It should be used alongside other journal evaluation factors rather than as the sole criterion. CiteScore is not appropriate for evaluating individual papers or researchers, and cross disciplinary comparisons should use percentiles rather than raw scores.

A common error is treating CiteScore as interchangeable with the journal impact factor; the two metrics use different databases, different citation windows, and different document type definitions, and their values are not directly comparable. Another mistake is using CiteScore to evaluate the quality of individual articles, when it describes the average performance of a journal's entire output.

Researchers sometimes also compare raw CiteScore values across disciplines without normalization. A CiteScore of 5.0 may represent excellent performance in one field but average performance in another, depending on disciplinary citation norms. CiteScore percentiles address this issue by ranking journals within their subject categories. Finally, assuming that a high CiteScore guarantees fast publication, a good match for a specific manuscript, or rigorous peer review is an error; CiteScore measures citation impact, not these other dimensions of journal quality.

A health economics researcher compares CiteScores for three target journals that share a Scopus subject category (the figures are illustrative). Journal A has a CiteScore of 7.2 (85th percentile in that category), Journal B has a CiteScore of 8.5 (91st percentile), and Journal C has a CiteScore of 4.8 (72nd percentile). The researcher uses these data alongside information about journal scope, audience, and turnaround time to select Journal B as the primary target, noting that its broader readership aligns with the manuscript's topic.