Submitting to Science.

01 About Science

Science, published by the American Association for the Advancement of Science (AAAS), is one of the most widely read and cited multidisciplinary research journals in the world. With a catalog of 386,855 works and a cumulative cited-by count of 22,615,221, the journal's reach spans virtually every corner of the scientific enterprise. Its top subject areas, as reflected in its published output, include medicine, astronomy and astrophysics, physics and astronomy, social sciences, economics and econometrics, history and philosophy of science, and the arts and humanities. This breadth is not incidental; Science has long positioned itself as a venue for work that transcends disciplinary boundaries and carries implications beyond a single field.

The journal's h-index of 1,746 and a two-year mean citedness of 18.43 reflect the sustained influence of its publications across decades. Science is not open access by default, though authors may have options for open access publication depending on their institutional agreements. Researchers across career stages read Science, from graduate students tracking the frontier of their field to senior investigators watching for paradigm-shifting results. The journal publishes original research articles, reviews, technical comments, and policy-relevant perspectives, making it a destination for work that is both scientifically rigorous and broadly significant.

02 Reproducibility and Data Standards

Science publishes work across an extraordinary range of disciplines, from environmental chemistry and computational biology to materials science and astrophysics, and the expectations around reproducibility reflect that breadth. Regardless of field, reviewers expect manuscripts to provide enough methodological detail that an independent researcher could replicate the study or, at minimum, critically evaluate how the conclusions were reached.

For experimental work, this means specifying reagent sources, instrument models, measurement protocols, and calibration procedures. For computational studies, which are increasingly prominent across every discipline Science covers, code availability, version-pinned software environments, and dataset documentation are becoming standard expectations. The shift toward open data and open code is not merely a cultural preference; it is a structural safeguard against the irreproducibility that has eroded confidence in published findings across multiple fields.

Authors should also be aware of field-specific reporting frameworks where applicable: CONSORT for clinical trials, ARRIVE for animal studies, and FAIR principles (Findable, Accessible, Interoperable, Reusable) for data management. Even when Science does not mandate a specific checklist, manuscripts that demonstrate awareness of these standards signal methodological fluency to editors and reviewers.

03 Common Methodology Concerns

Reviewers evaluating manuscripts for Science, across its full disciplinary range, consistently flag a set of interrelated methodological issues. Understanding these concerns before submission gives authors a meaningful opportunity to address them proactively.

One of the most common issues is insufficient detail in experimental or computational methods. Reviewers expect enough specificity to evaluate whether the conclusions are supported: for bench science, this means complete descriptions of materials, controls, and measurement protocols; for computational work, it means specifying model architectures, training procedures, hyperparameters, and hardware. Vague methods sections are a persistent reason manuscripts are returned before review even begins.

Sample size justification is another frequent concern. Whether the study involves field samples, cell lines, animal models, or computational benchmarks, reviewers want to see a rationale for why the chosen sample or dataset size is adequate to support the claimed conclusions. For statistical analyses, this often means a power analysis; for computational experiments, it means demonstrating that results are stable across multiple runs or random seeds.

The handling of controls and negative results also receives close scrutiny. Reviewers at Science expect appropriate positive and negative controls for experimental work, ablation studies or sensitivity analyses for computational and modeling studies, and honest reporting of conditions under which the method does not perform well. Finally, the distinction between confirmatory and exploratory findings must be clear. Claims supported by pre-planned analyses carry more weight than post-hoc observations, and conflating the two is a reliable path to major revision requests.

04 Recent Representative Work

The following papers, published in Science in 2025, illustrate the journal's thematic range, from environmental science and computational biology to materials chemistry and energy technology:

These papers reflect the kind of work Science tends to publish: studies with broad scientific significance, methodological rigor, and implications that extend beyond a single discipline or application domain.

05 Pre-Submission Checklist

06 How PeerPanel Reviews Your Manuscript

Before your manuscript reaches Science's editorial team, PeerPanel runs it past five specialist agents, each focused on a distinct dimension of manuscript quality.

07 Where to Find Author Guidelines

The official author guidelines for Science are available on the journal's homepage at sciencemag.org. Authors should consult these guidelines directly for current requirements on manuscript formatting, word and figure limits, supplementary material policies, and submission procedures. Guidelines are updated periodically, and the journal's own instructions take precedence over any third-party summary. PeerPanel is a pre-submission tool; it does not replace reading the official instructions, and authors are responsible for ensuring their submission meets all journal requirements before uploading.