First and last authors more likely to be men in leading science journals

In some ways, women have made incredible gains in science, technology, engineering and mathematics (STEM). In 2022, they represented 41% of all active researchers, globally — up from 28% in 2001 — and 44% of new US science and engineering doctorates in 2023.

However, an analysis of first and last authorship in journals tracked by Nature Index reveals a persistent gender imbalance in these markers of scientific contribution and authority, indicating that recognition in top-tier academic publishing has not kept pace with women’s growing presence in research.

Among the natural-sciences journals tracked by the index, women represented 29% of first-author positions and 17% of last-author positions in 2025. These figures have improved only slightly over the past decade (see ‘Slow progress’), up from 28% of first-author positions and 15% of last-author positions in 2015, and are nowhere near gender parity — having women make up 40–60% of these authorships.

The fact that women make up less than one-third of first authors and less than one-fifth of last authors in this data set is troubling because of the status that is baked into these positions, says Cassidy Sugimoto, an information scientist at the Georgia Institute of Technology in Atlanta.

The first-author position is generally given to the researcher who did most of the work, and the last-author position tends to be reserved for the principal investigator. For junior and senior researchers, first and last authorships “are the coin of the realm in science”, Sugimoto says, and can mean career opportunities, grants and “other kinds of accolades in the academic market”.

Compared with the overall proportion of female authorships in the natural sciences in 2025 (31%), the percentage of women in last-author positions is especially low. A similar pattern is seen in the health sciences. Although women accounted for 41% of authorships in health-sciences papers in 2025 and made up 44% of first authors, they held only 31% of last authorships.

Because Nature Index tracks only a selected set of high‑impact journals, the findings point to persistent barriers for women at the highest levels of academic publishing.

Accounting for the Index’s focus on top‑tier journals is “super important”, says Curt Rice, founder of Publishing Unlocked, an Oslo‑based initiative that supports early‑career researchers. “It’s not so much about who is doing the research,” he says, “but about who is being recognized at the top of the publication hierarchy.”

A closer look

For the Nature Index analysis, gender is inferred by an algorithm that uses name–gender associations in an author’s most likely country of origin. The first results using this approach were released last year. The analysis has now been expanded to include authorship-position data at a subject, topic and journal level.

Several limitations should be considered when interpreting trends in these data. Gender is inferred rather than self‑reported, so the findings should be treated as indicative rather than definitive. The data are incomplete in some countries, such as China and Singapore, because the model could not reliably determine gender for researchers in those locations.

Furthermore, some research fields do not follow the authorship convention set by most other disciplines. Economics, mathematics and linguistic journals, for example, tend to list authors in alphabetical order, and some fields — such as high-energy physics — have hundreds of authors on a single paper, which waters down the weight carried by first and last authorship. These fields were not excluded from the analysis, so the interpretations need to be viewed with caution. In the past, some countries, such as China, have placed a higher value on first authorship and provided economic incentives for being listed first¹, which also skews the dynamics.

It’s also important to note that the Nature Index analysis counts authorship instances rather than individual authors, which means that a scientist whose name appears on five papers would count as five authorships, rather than one, for example. This approach amplifies highly networked individuals and could be affecting the numbers, Rice says. “If men dominate there — which they do — then the metric itself reflects accumulated advantage.”

Some of the strongest trends can be seen at a subject level, at which large gender disparities persist in authorship positions. In the physical sciences, women accounted for the smallest share of authorships in 2025, making up 20% of first authors and 13% of last authors — up from 16% and 11%, respectively, in 2015. Chemistry shows a similar pattern, although with slightly higher representation: women made up 28% of first authorships and 14% of last authorships in 2025.

Even in fields that have relatively high representation of women, progress has been modest. In the biological sciences, women accounted for 39% of authorships in 2025, overall, and 42% of first authors and 24% of last authors. This represents only a small increase from 2015, when women held 39% of first and 21% of last authorships.

At a topic level, the largest gender gaps are observed in fields with high overall female participation (see ‘Gender gap’). For example, in paediatric research in 2025, 62% of first authors but only 45% of last authors were women. In the clinical sciences, women had 44% of first and 27% of last authorships. This pattern reflects strong female representation at junior career stages that is not seen in senior authorship roles.

By contrast, fields dominated by men show smaller gaps, mainly because female representation is low across career stages. In quantum physics, for example, 15% of first and 9% of last authorships were women in 2025.

Behind the curtain

The Nature Index analysis does not follow individual researchers over time, nor does it capture differences in submission patterns, editorial decision-making, funding access or career interruptions that might shape who appears in top-tier journals. As a result, the findings indicate persistent disparities in recognition, but cannot, on their own, identify where along the publication pipeline those disparities emerge or accumulate.

Even so, the persistence of the gap over several years challenges the assumption that increasing women’s participation at early-career stages will, by itself, deliver parity at senior levels, says Rice. Instead, the pattern points to obstacles for women during the transition from junior to senior roles. Social and structural factors might be affecting men and women differently.

Track gender ratios in research to keep countries, institutions and publishers accountable

“The assumption that today’s first authors will be tomorrow’s last authors leaves out a whole lot of stuff that happens in between those stages, and that plays out differently for men and women,” says Rice.

The findings chime with what the research has been showing. For years, researchers have been calling attention to the fact that women are more likely than men to be middle-position authors. Even in studies that list two or more researchers who contributed equally and share the first-author position, men are usually listed before women².

Men are also more likely to have a first-authorship position earlier and more often in their careers, and some evidence suggests that, at a population level, men are more likely than women to appear as authors in high‑impact journals. A 2023 paper, for example, found that the higher a journal’s impact factor (a metric based on how often a journal’s articles are cited and used as one measure of its influence), the less likely a woman is to appear as first or last author³.

Another paper published in January estimated that biomedical and life-sciences articles with women as the first or corresponding author spend 7—15% longer under review than do ones authored by men⁴.

These trends reflect deeper inequities and unconscious biases, says Elizabeth Pollitzer, co-founder and director of Portia, a London-based non-profit organization that focuses on gender-related issues in STEM. “It’s frequently said that men are assessed on potential, and women are assessed on actual achievement,” she says.