Why the Same Citation Search Gives Different Results on Crossref and OpenAlex
We tested citation verification against both databases and found surprising differences. Here's what each one is good at, where they fall short, and why Cite Checker uses both.
Same paper, different search results
While building Cite Checker, we kept running into a pattern: search for a citation on Crossref, get nothing. Search for the same citation on OpenAlex, and there it is. Sometimes the reverse happened too.
A 2020 Japanese institutional repository paper? Crossref returned zero results (no DOI assigned), but OpenAlex found it by title. Meanwhile, a Springer Nature journal article? Crossref had more accurate metadata than OpenAlex.
This is why Cite Checker queries both databases. Here's the reasoning.
Crossref: the publisher's official record
Crossref is a non-profit founded in 2000 that manages DOIs for scholarly publications. As of 2025, it holds over 150 million records.
The key detail is how data gets in. When a publisher releases a journal article, the publisher themselves registers the metadata with Crossref. This means the title, authors, year, and volume information carries publisher authority. High accuracy, but if a publisher doesn't register, the data simply doesn't exist.
Where Crossref excels
- Major journal articles (Elsevier, Springer, Wiley, IEEE, etc.)
- DOI verification (it's the DOI registry, after all)
- Precise bibliographic metadata
- Citation link data (via I4OC — Initiative for Open Citations)
What Crossref misses
- Papers without DOIs (older works, some institutional repositories)
- Preprints (many arXiv and bioRxiv papers lack Crossref DOIs)
- Book chapters (registration is inconsistent)
- Smaller journals from non-English-speaking countries
OpenAlex: casting a wider net
OpenAlex launched in 2022 as an open scholarly database, succeeding the discontinued Microsoft Academic Graph. It indexes over 250 million records.
The biggest difference from Crossref is how it collects data. OpenAlex pulls from multiple sources automatically — including Crossref itself, plus journal websites, institutional repositories, and preprint servers. No DOI required.
Where OpenAlex excels
- Finding papers without DOIs
- Preprints (arXiv, bioRxiv, etc.)
- Author disambiguation and institutional affiliation data
- Research field classification and trend analysis
- Non-English literature coverage
Where OpenAlex falls short
- Auto-collected metadata can be less accurate than Crossref's publisher-registered data
- Author name matching sometimes produces false positives
- Started in 2022, so some historical data has gaps
- API responses tend to be slightly slower
Side-by-side comparison
Feature Crossref OpenAlex Founded 2000 2022 Records 150M+ 250M+ Data source Publisher-registered Auto-collected from multiple sources DOI required Yes No License Free API CC0 (fully open) Author info Basic Detailed (ORCID linked) API speed Fast, stable Sometimes slower Why Cite Checker uses both — with real numbers
Crossref first, for accuracy
Publisher-registered data means precise title and author matching. For citation verification, the reference database needs to be accurate — otherwise you can't tell whether a mismatch is the user's error or the database's error.
Cite Checker's confidence score is a weighted average: title similarity (50%), author match (30%), publication year (20%). The more accurate the reference data, the more meaningful that score is. That's why Crossref goes first.
OpenAlex as fallback, for coverage
When Crossref returns a confidence score below 80%, Cite Checker also queries OpenAlex.
This matters in two specific cases. First, papers without DOIs — they literally don't exist in Crossref. Second, citations where the author name or title spelling differs significantly from the Crossref record. OpenAlex's broader collection can catch what Crossref misses.
In our testing (8 academic papers, 475 total citations), about 3% of citations were only found via the OpenAlex fallback. That's not a huge number, but those citations going from "Not Found" to "Found" makes a real difference for users trying to distinguish real references from fake ones.
Practical tips for researchers
Checking your own citation count
You can query the Crossref API with your paper's DOI to get citation counts. But Crossref only tracks citations from I4OC-participating publishers, so the count will be lower than Google Scholar. OpenAlex generally has more complete citation relationship data.
Making literature reviews more comprehensive
For systematic literature reviews, try Google Scholar + OpenAlex together. OpenAlex's Concepts feature lets you find related papers by research topic, which is useful for discovering work you might have missed.
Verifying preprint citations
arXiv preprints often aren't registered in Crossref. For preprint verification, search OpenAlex or Semantic Scholar instead. Both index preprint servers comprehensively.
The bottom line
Crossref and OpenAlex aren't competitors — they're complementary. Crossref brings accuracy; OpenAlex brings breadth. Cite Checker combines both because neither one alone catches everything.
Understanding what each database does well (and where it has blind spots) is useful beyond citation checking. It helps with literature search, tracking research impact, and spotting gaps in your reference list.