
At 9 a.m. on a Tuesday, the health desk of a national newspaper holds its news meeting. The editor has fifteen minutes, forty emails, and room for three science stories. Your press release is one of those forty emails. If the headline does not carry the news, if the first paragraph does not answer the reader's questions, the email is closed and the study never reaches the public.
Most research papers are read by a small circle of specialists, yet the findings often matter to a much wider audience: patients, farmers, teachers, policymakers, and the public who funded the work. A press release is the bridge between the two. This guide explains how to write one, from the headline to the final checklist, using a single worked example: a veterinary study on milk yield that you will see built into a headline, a lead, a quote, and a full release.
Not every paper needs a press release, and sending one for the wrong paper wastes a journalist's time and your credibility. Ask three questions before you start. Is the finding new, or does it confirm what was already known? Does it affect people outside your field? Can the result be explained in plain language without losing its meaning? If the answer to all three is yes, write the release.
Take the worked example for this guide. Dr Sana Iqbal, at the University of Veterinary and Animal Sciences in Lahore, studied 240 dairy farms over two summers and found that a simple change in afternoon feeding times raised milk yield by an average of 1.8 litres per cow per day. The finding is new, it affects farmers and milk prices, and it can be explained in one sentence.
The headline is the only part of your release that most journalists will certainly read, so it must carry the news on its own. State the finding, not the topic: name what changed, who it affects, and why it matters. A weak headline for Dr Iqbal's study would read: "New study on dairy cattle feeding schedules published." It says what the study was about, and nothing about what it found. A strong headline reads: "Changing afternoon feeding times raises milk yield by nearly two litres a cow a day, study of 240 farms finds."
Keep the headline short enough to read in one breath, and avoid jargon a non-specialist would not recognise. Do not use abbreviations, and do not promise more than the paper proves.
The first paragraph, called the lead, must answer the questions a reader asks first: what happened, who did it, where, when, and why it matters. Here is the lead for Dr Iqbal's study: "Dairy farmers can raise milk yield by an average of 1.8 litres per cow per day by moving afternoon feeding to two hours earlier, a study of 240 farms in Punjab has found. Researchers at the University of Veterinary and Animal Sciences, Lahore, followed the farms over two summers and found the gain held across small and large herds. The change costs farmers nothing, and the authors say it could be adopted immediately."
Write the lead after you have drafted the rest of the release, when you know exactly what the story is. Put the strongest, most newsworthy fact first, then the context that makes it understandable. Avoid background history in the lead; it belongs later, after the reader knows why they should care. If the lead only sets the scene and saves the finding for later, most readers will never reach it.
After the lead, the body explains the study in the order a reader needs it, most important first. For Dr Iqbal's study, the order is: what the researchers did, in two sentences; what they found, with the key numbers; what it means for farmers; and finally the background, including how the study was funded and where it is published. This is the inverted pyramid: the most important information first, the background last, because editors cut from the bottom.
Keep the method short and plain: one or two sentences on how the study was done, enough for the reader to judge that it was careful, without the detail that belongs in the paper itself. For example: "The team visited each farm monthly, recorded feed times and milk yield, and compared farms that changed feeding times with farms that did not."
A good quote sounds like something a person would actually say, not like a sentence from a paper. It should carry the human meaning of the finding: what surprised the researcher, what it could change, or why the work mattered to them. A weak quote for this study would read: "Our results demonstrate a statistically significant increase in milk yield associated with modified feeding schedules." No journalist will use that sentence, because it says nothing a human would say. A strong quote reads: "We expected a small gain, and we were surprised by how consistent it was. On farm after farm, large and small, the same change gave the same result. A farmer can try this tomorrow, and it costs nothing."
Write the quote in the researcher's own voice, then ask them to approve it word for word, because their name will be attached to it. Keep it to two or three sentences at most, and avoid repeating the numbers already given in the text. One strong, honest quote is worth more than three cautious ones that say the same thing as the lead. If the researcher is uncomfortable with a quote you drafted, ask them what surprised them most about the study: the best quotes usually come from that question.
Journalists trust numbers more than adjectives, so give them the figures that matter: how many people, animals, or samples were studied, how large the effect was, and how it compares with what came before. In Dr Iqbal's release, the numbers are: 240 farms, two summers, 6,400 cows in total, an average gain of 1.8 litres per cow per day, and a gain that appeared on 212 of the 240 farms. A percentage without the underlying number is weak, so the release gives the litres and the number of farms, not only a percentage.
Be honest about uncertainty, because a release that states its limits clearly is more credible, not less. The release for this study says plainly that the gain was smaller on farms where cows were already fed early, and that the study covered one region and two summers, so the result may differ elsewhere.
An embargo is an agreement that journalists may read a study before publication but may not report it until a stated date and time. Journals use embargoes so that reporters have time to prepare accurate stories and the news does not appear before the paper itself. Here is how the timeline works for a typical paper. The paper is accepted on 3 March. The journal sets the embargo to lift on 10 March at 10:00 GMT. The press office sends the release to journalists on 7 March, marked clearly at the top: "EMBARGOED UNTIL 10 MARCH, 10:00 GMT".
Never send a release under embargo without stating the embargo clearly at the top, in bold, with the exact date and time. If you are unsure whether your paper is under embargo, ask the journal before you contact any journalist. Breaking an embargo can damage your relationship with the journal and with the reporters who trusted you.
Send the release to people who cover your field, not to every journalist you can find. A health reporter, a science correspondent, and an editor at a specialist outlet are far more likely to use your story than a general news desk. Build the list in four steps. First, search for who has covered similar studies in the last year. Second, add the agriculture or science correspondent of the two largest newspapers in your country. Third, add one specialist outlet, such as a farming magazine. Fourth, check that each name is still at that outlet.
Use a clear subject line that repeats the headline, and paste the release into the body of the email rather than attaching it as a file, because attachments often go unopened. Include your name, role, and phone number at the end, and offer the full paper and an interview. A release that is easy to open, read, and act on is far more likely to become a story.
Your university or institute almost certainly has a press office, and it should be your first partner, not an afterthought. Press officers know which journalists cover your field, how to time a release around a journal's embargo, and how to shape a story for a general audience. Contact the press office as soon as the paper is accepted, not the day before publication.
Share the paper, the key findings, and the names of the researchers available for interview, and tell the press office the embargo date on the first call. The press office works for the institution, so it may ask for changes: a different headline, a shorter quote, or a change to the timing. Its aim is the same as yours, so treat its edits as help.
A press release for a research paper should fit on one page, or two at most. One page forces you to keep only what matters. Count the words before you send: a release of 350 to 450 words is about right for a single study.
End the release with a short boilerplate: two or three sentences about your institution or publisher, and a line saying where the paper is published. Add the word ENDS or three hash marks after the main text so the reader knows the story is finished, then put the media contact details below it: name, role, phone, and email. This format is standard, and journalists recognise it instantly.
If a journalist does not reply, wait at least a day before following up, and follow up only once. A short, polite message that adds something new is far more effective than repeating the original release: "The paper is now published, and Dr Iqbal is available for interview between 10 a.m. and noon tomorrow." Do not call a journalist repeatedly.
Run this checklist before you send. The headline carries the news, and the lead answers what, who, where, when, and why. Every number has been checked against the paper, and every name is spelled correctly. The embargo date and time have been confirmed with the journal, and the embargo is stated at the top. The quoted researcher has approved their words, the press office has seen the release if your institution requires it, and the media contact details are at the bottom.
| Part | Purpose | Suggested length |
|---|---|---|
| Headline | Carry the news in one line | Under 15 words |
| Lead paragraph | Answer what, who, where, when, why | 3 to 5 lines |
| Body | Method, findings, and meaning | 3 to 5 short paragraphs |
| Quote | Add human meaning and judgement | 2 to 3 sentences |
| Boilerplate | Say who you are and where to find the paper | 2 to 3 sentences |
| Contact details | Let journalists reach you fast | Name, role, phone, email |
These authoritative external resources explain the standards and tools mentioned in this guide:
When should I send a press release?
Send it when the paper is accepted and you know the publication date, so the story can appear on the day the paper is published. If the journal sets an embargo, respect it exactly. Sending one for a paper still under review is too early, and sending one weeks after publication is usually too late.
How long should a press release be?
One page is ideal, and two pages is the most a journalist will read. The headline, lead, evidence, one quote, and contact details should fit comfortably in that space. If your draft runs longer, the story has not been focused yet, and cutting it yourself is better than letting an editor cut it for you.
What is an embargo?
An embargo is an agreement that journalists may read a study early but may not publish anything about it until a stated date and time. Journals use embargoes so reporters can prepare accurate stories. The embargo date is set by the journal, and breaking it can damage your relationship with the journal and with reporters.
Do I need my university press office?
In most cases, yes. Press officers know the journalists who cover your field, understand embargoes, and have distribution lists you do not. Contact them as soon as the paper is accepted, share the paper and the key findings, and let them shape the release with you. It is a partnership, not a formality.
Can I send a press release for a preprint?
Only if the journal that will publish the work allows it, and only with care. Some journals treat publicity before acceptance as a problem, and a preprint is not peer-reviewed, so any release must say so plainly. Check the journal's policy first, and never present preprint findings as if they had already passed peer review.
What if no journalist replies?
That is normal, and it is not a failure. News desks receive dozens of releases a day, and a story may simply not fit that outlet on that day. Follow up once, politely, with something new to add, then move on. Keep the relationship warm, because a journalist who trusts you will read your next release.
Submit your manuscript ← Back to GAP Insights
Useful publishers, databases, funders, tools and institutions for researchers: