Extended definition
A cover letter is the short document — typically one to two pages — accompanying manuscript submission to a journal, nominally addressed to the editor-in-chief or handling editor. Its function is to articulate, in direct prose, three central elements: (1) work relevance for the specific journal (not general scientific relevance, but specific fit with scope, audience, and editorial tradition); (2) standard editorial declarations required by most journals — originality, no parallel submission elsewhere, conformity with ICMJE authorship criteria, disclosure of conflicts of interest, ethical compliance in human/animal research; (3) highlights of what makes the work a substantive contribution to the field. Bourne (2005, PLoS Computational Biology) and Provenzale (2007, AJR) offered canonical references in their “Ten Simple Rules” series on academic editorial practice. The letter often operates as an initial editorial filter: experienced editors read the cover letter first to decide whether the manuscript merits peer review or should be desk-rejected. A weak or misaligned letter can compromise an otherwise technically sound manuscript.
When it applies
A cover letter applies to every journal submission. Modern submission systems (ScholarOne Manuscripts, Editorial Manager, eJournalPress) require separate upload of the letter. It applies to resubmissions after rejection: the letter should mention prior history if relevant (rejection at another journal need not be declared; prior rejection at the current journal with invitation to resubmit should be). It applies to fast-track or expedited review submissions: justifying urgency is central. It applies to submissions transferred between journals of the same group (Nature Portfolio, Springer cascade): use the opportunity to adjust argumentation for the new journal. Modern best practice: personalized letter for the specific journal, not a generic template.
When it does not apply
It does not apply at top-tier computer science conferences (NeurIPS, ICML, ACL) — submission is via OpenReview or similar, without cover letter; relevance argument goes into the paper itself. It does not apply as space for substantive scientific content: methodological details belong in the manuscript, not the letter. It does not apply as justification for relaxing editorial criteria: the letter does not replace methodological rigor. It does not apply as a vehicle to claim temporal priority over competitors: priority is established by dated preprint and publication DOI, not by letter assertion. It does not apply at some journals that replaced cover letters with structured submission fields (significance statement, novelty statement) — check specific editorial requirements.
Applications by field
— Health and biomedical sciences: Lancet, NEJM, JAMA require detailed letters with ICMJE-standard declarations and justification of clinical relevance. — Social sciences: letter typically discusses theoretical contribution and implications; fit with the journal’s editorial tradition. — Computer science and engineering (journals): similar pattern at IEEE Transactions, ACM TOG, Nature Communications; conferences use a different mechanism. — Humanities: letter may discuss interpretive contribution and engagement with the journal’s specific debates; less standardization than in STEM.
Common pitfalls
The first pitfall is a generic template applied to multiple journals without personalization: experienced editors detect it immediately, signaling lack of care. The second is overstating relevance with superlatives (“first”, “definitive”, “revolutionary”) when the work is an increment — credibility compromises subsequent manuscript reading. The third is omitting standard editorial declarations: the editor may reject for administrative incompleteness before even evaluating content. The fourth is failing to articulate fit with the journal: a letter that could have been sent to any journal in the field is weak; journals have specific editorial identities. The fifth is failing to correct the editor’s or journal’s name after template reuse: a letter-by-letter error like “Dear Editor of Lancet” sent to BMJ is near-automatic desk-reject.