Rank of Nature and Science among all journals, measure by measure
Where each stands among all journals on each measure, in its latest ranked year: further right is a higher rank, and the figures are the "top x%" each one holds. Publications, citations and h-index grow with size, so a larger body will sit further right on those.
Disciplines where Nature or Science holds the larger share of its output
Where do they differ? On the fixed map of disciplines, each circle marks a discipline taking a larger share of the output of the entity in its colour. Area is the gap between the two shares, so size doesn't count: a small specialist can out-weigh a large generalist.

Add a third: Proceedings of the National Academy of Sciences of the United States of America · Science Advances · Nature Astronomy · Nature Communications · Eos · Symposium - International Astronomical Union

Radar chart comparing Nature and Science on speed, impact, depth, cited share, self-citation and retractions
Further out is better on every spoke. Each axis is scaled to the better of these two journals rather than to the whole corpus, so one of them reaches the rim on each axis and the shape shows how the other compares. The scale is logarithmic: impact and h-index run over orders of magnitude, and on a linear axis a strong specialist journal beside a giant would draw as a dot at the centre.
Choose Nature for reach.
Choose Science for cleaner citation metrics and a paper you intend to publish open access.
SpeedNatureScienceBetter
Publication speedMedian days from submission to publication over the last three years.355 days——
ReachNatureScienceBetter
2-year impactCitations in the latest impact year to eligible publications from the previous two years, divided by those publications.54.9543.34Nature
5-year impactCitations in the latest impact year to eligible publications from the previous five years, divided by those publications.57.4849.84Nature
h-indexDepth of the journal's cited record rather than its recent average.1,7621,698comparable
Score D tierWhere the journal's prestige-weighted citation ratio places it among all journals. Tier 1 is the top 5%.11comparable
Risk to your paperNatureScienceBetter
Uncited shareHow much of the journal's output is never cited at all — the chance your paper is not read.16.7%38.5%Nature
Retraction notices per 1,000 articlesA high rate says something about what gets through review.0.390.21Science
Metric integrityNatureScienceBetter
Self-citation share of 2-year impactHow much of the headline impact figure the journal generates by citing itself.1.3%0.9%Science
Journal self-citation rateThe same across the whole record rather than one window.1.2%0.9%Science
Author self-citation rateCitations whose citing paper shares an author with the paper cited.5.8%5.8%comparable
Scale and scopeNatureScienceBetter
Articles per yearRecent output. A larger journal is not better or worse, but it is a different place to publish.1,1681,007—
Subject breadthDisciplines the journal is classified under.———
National quality levelLevel under the national research-assessment classification; 2 and 3 are the quality tiers.33comparable
Data available to downloadShare of publications, reviews excluded, whose data can be downloaded rather than only requested.43.6%18.3%—
If you publish open accessNatureScienceBetter
Open-access shareShare of publications carrying an open licence. Counted only where a licence was actually found, so it is a floor. Openness usually carries a publication charge, so this describes the journal rather than ranking it.2.1%0.1%—
Open-access citation advantageHow much more an open article is cited than a paywalled one published in the same year in the same journal. Above 1 means opening a paper here is associated with more citations. It is an association, not an effect: authors choose which work to open, and that choice is not independent of how citable the work already was.1.462.18Science

Author countries

The union of each journal's top ten. Shares use papers with known country affiliations and overlap. Geography does not contribute to the comparison verdict.

CountryNatureShareScienceShare
United States2,52468.5%1,50464.7%
China93925.5%55924.1%
United Kingdom93125.3%46920.2%
Germany83722.7%44018.9%
France44512.1%25911.1%
Switzerland41911.4%2189.4%
Canada3429.3%2068.9%
Japan3409.2%1918.2%
Spain3058.3%1275.5%
Australia2988.1%1697.3%
Netherlands2927.9%1345.8%

Nature: country coverage and screening

Insufficient field reference data (30 established peers required). A Watch requires at least 30 recent papers from the country, a share at least 5 percentage points above the peer 95th percentile, and a rise of at least 15 points over dated papers before 2024. Both periods need 100 country-attributed papers and 80% country coverage, with coverage differing by no more than 10 points. These are screening thresholds, not a finding about journal quality. Stable national concentration is context.

2024–2026: 3,685 of 3,781 papers have country affiliations (97.5%). Shares use these papers and their 238,211 citations; countries overlap.

Dated papers before 2024: 94,602 of 163,684 papers have country affiliations (57.8%). Shares use these papers and their 21,765,706 citations; countries overlap.

Recent top 10 by articles: country shares and observed peer reference
CountryRecent papersRecent shareEarlier shareChangePeer medianPeer 95th percentileScreen
United States2,52468.5%56.7%+11.8 pp——Insufficient data
China93925.5%2.5%+22.9 pp——Insufficient data
United Kingdom93125.3%20.0%+5.3 pp——Insufficient data
Germany83722.7%6.4%+16.3 pp——Insufficient data
France44512.1%5.5%+6.5 pp——Insufficient data
Switzerland41911.4%3.4%+8.0 pp——Insufficient data
Canada3429.3%5.3%+4.0 pp——Insufficient data
Japan3409.2%4.9%+4.3 pp——Insufficient data
Spain3058.3%1.7%+6.6 pp——Insufficient data
Australia2988.1%4.9%+3.1 pp——Insufficient data

Science: country coverage and screening

Insufficient field reference data (30 established peers required). A Watch requires at least 30 recent papers from the country, a share at least 5 percentage points above the peer 95th percentile, and a rise of at least 15 points over dated papers before 2024. Both periods need 100 country-attributed papers and 80% country coverage, with coverage differing by no more than 10 points. These are screening thresholds, not a finding about journal quality. Stable national concentration is context.

2024–2026: 2,324 of 2,594 papers have country affiliations (89.6%). Shares use these papers and their 111,505 citations; countries overlap.

Dated papers before 2024: 64,317 of 135,717 papers have country affiliations (47.4%). Shares use these papers and their 18,430,395 citations; countries overlap.

Recent top 10 by articles: country shares and observed peer reference
CountryRecent papersRecent shareEarlier shareChangePeer medianPeer 95th percentileScreen
United States1,50464.7%82.0%-17.3 pp——Insufficient data
China55924.1%2.5%+21.6 pp——Insufficient data
United Kingdom46920.2%7.9%+12.3 pp——Insufficient data
Germany44018.9%6.2%+12.7 pp——Insufficient data
France25911.1%4.2%+6.9 pp——Insufficient data
Switzerland2189.4%2.7%+6.7 pp——Insufficient data
Canada2068.9%4.4%+4.4 pp——Insufficient data
Japan1918.2%3.6%+4.6 pp——Insufficient data
Australia1697.3%2.2%+5.0 pp——Insufficient data
Netherlands1345.8%2.0%+3.7 pp——Insufficient data
This compares the indexed record only. It describes how each journal has behaved, not the quality of any individual paper, and says nothing about whether either would accept yours. Full detail: Nature health check · Science health check · compare a different pair.
Download this comparison:JSONCSV