Scopus

Scopus is Elsevier’s proprietary abstract-and-citation database of peer-reviewed literature — journal articles, conference proceedings, books, and book chapters — with each indexed record’s own reference list captured so citation links can be traced between records. It is a subscription product: there is no free public search interface, unlike OpenAlex. Elsevier launched Scopus publicly on 3 November 2004 at the Science Museum in London, positioning it as a direct competitor to Thomson Reuters’ (now Clarivate’s) Web of Science and its Science Citation Index. What makes a record “in Scopus” is inclusion in Elsevier’s own title-selection and re-evaluation process (the Content Selection and Advisory Board reviews journals against criteria such as peer-review policy, publication ethics, and international editorial diversity) — being indexed is an editorial decision by the database operator, not an automatic consequence of a journal existing or being open access.

Coverage and scope

As of early 2025, Elsevier reports Scopus covering more than 100 million records drawn from over 28,000 active, currently-indexed journals and conference proceedings from more than 8,000 publishers, with selective historical coverage reaching back to 1788 for some titles, though the great majority of full indexing depth begins around 1970 onward. Content spans more than 330 subject-classification categories across physical sciences, life sciences, health sciences, and social sciences and humanities.

Scopus’s coverage overlaps heavily with, but is not identical to, Web of Science’s Core Collection — the two databases use different journal-selection criteria and cover a different (though substantially overlapping) set of titles, which is why the same paper can carry a different citation count depending on which database counted it. For a full side-by-side comparison of ownership, access model, reported scale, document types, and which database a research office would reach for which task, see Scopus vs Web of Science vs OpenAlex.

Scopus’s bibliometric indicators: CiteScore, SJR, and SNIP

Scopus data underlies three journal-level citation metrics that a research office will routinely encounter in tenure/promotion files, journal-selection guidance, and research-assessment reporting. They are related but not identical in either ownership or calculation:

• CiteScore is Elsevier’s own metric, introduced in 2016 as the Scopus-based counterpart to Clarivate’s Journal Impact Factor. It is calculated directly by Elsevier from Scopus data and published free at scopus.com/sources without requiring a subscription to search the underlying database. See the CiteScore entry for its exact calculation window and methodology history.
• SJR (SCImago Journal Rank) is not an Elsevier product — it is produced independently by the SCImago Research Group (a Spanish research group associated with the Consejo Superior de Investigaciones Científicas and the University of Granada), using Scopus data under license. SJR weights each incoming citation by the prestige of the citing journal, using an algorithm related to Google’s PageRank, so a citation from a highly-cited source counts for more than one from a marginal source. It is published free via the SCImago Journal & Country Rank portal (scimagojr.com) and also surfaced on Scopus source pages.
• SNIP (Source-Normalized Impact per Paper) is likewise not calculated by Elsevier — it was developed by Henk Moed and is calculated from Scopus data by the Centre for Science and Technology Studies (CWTS) at Leiden University. SNIP corrects for field-level differences in citation behavior (fields where citations are less common get more weight per citation), which makes it, of the three, the one specifically designed for comparing journals across different subject areas rather than within a single field.

All three are journal-level indicators and inherit the same limitations any journal-level metric has when applied to an individual article or an individual researcher’s output — a caution reflected in the San Francisco Declaration on Research Assessment (DORA)‘s position against using journal-level metrics as a proxy for individual research quality. See Journal Impact Factor and Quartile Rankings for how Scopus-based quartile placement (using CiteScore) compares to the Web of Science equivalent, and Field-Weighted Citation Impact (FWCI) for a related Scopus-derived indicator used at the author/institution level rather than the journal level.

Role in research administration

Beyond journal-level metrics, Scopus assigns its own persistent identifiers — a Scopus Author ID for individual researchers and a Scopus Affiliation ID for institutions — used to disambiguate and aggregate an author’s or institution’s output within the database. These identifiers are commonly ingested by campus current research information systems (CRIS) such as Elements (Symplectic) and Pure to populate publication lists and compute institution-level bibliometric reporting, typically alongside or cross-walked against a researcher’s ORCID iD rather than replacing it. A research office relying on Scopus for institutional reporting should treat Scopus Author/Affiliation IDs as a Scopus-internal disambiguation layer, not an interoperable persistent identifier in the ORCID/ROR sense — they resolve only within Elsevier’s own ecosystem.

Worked examples

• A peer-reviewed journal article published in a title that has passed Elsevier’s Content Selection and Advisory Board review, with its full reference list captured, is a Scopus record — it will have a Scopus Author ID attached to each listed author and count toward that journal’s CiteScore, SJR, and SNIP for the relevant year(s).
• A preprint deposited only on a preprint server (with no accompanying journal publication) is generally not a Scopus record on its own — Scopus’s core indexing is built around peer-reviewed serial and proceedings content, not preprint repositories, though Elsevier has separately expanded some preprint coverage under specific arrangements; check current source coverage on scopus.com rather than assuming inclusion.

Counter-example

A journal listed in Web of Science’s Core Collection is not automatically in Scopus, and vice versa — the two databases run separate, independent selection processes with different criteria, so title overlap is substantial but not total. A citation count or h-index pulled from one database cannot be assumed to match a count pulled from the other for the same author or journal; see Scopus vs Web of Science vs OpenAlex for how the two selection processes differ.