No organization likes to feel the pressure of negative media coverage. Last year, when the Fred Hutchinson Cancer Research Center in Seattle was accused of mismanaging clinical trials, this bastion of the medical establishment found itself under an unwanted spotlight. Now the series of articles that made the allegations is a contender for the Pulitzer Prize, and those involved are reliving their media nightmare (see page 463).

Leading researchers and scientific bodies have rallied round the centre. This was bad journalism, say the critics. It seems like a familiar story: sloppy journalists misrepresent science.

How often does this happen? The cancer-centre story is a complicated and emotive one. But there is no doubt that science is, on occasion, badly covered in the media. Last week, for instance, one British newspaper ran a front-page story headlined “Scientists find Prozac 'link' to brain tumours”. The lead author of the research told Nature that the story was “scandalous” — there is no epidemiological link. His study merely showed that the antidepressant could, in the test tube, prevent certain tumour cells from being triggered to commit suicide by the neurotransmitter serotonin.

But such examples are the exception. In the United States, for instance, media coverage of climate change has been thorough, explaining the consensus view while noting dissenting voices.

In Britain, newspapers, including the one that carried the Prozac story, have produced some excellent science journalism. Much of the coverage of a recent public scare over the safety of the combined measles, mumps and rubella vaccine, for instance, stressed that most experts regard the vaccine to be safe, and pointing out the limitations of the small studies that have linked it to the development of autism. Even the Prozac story qualified the nature of the 'link' further down the piece. Take away the inflammatory headline, and the prominent positioning in the paper, and the story may have caused little fuss.

Nevertheless, many scientists are quick to attack the media when they believe they have been misrepresented. And at conferences addressing the public understanding of science, journalists are often portrayed as the root of the problem.

This knee-jerk reaction itself misrepresents the way in which stories enter the media — journalists are not the only players. University press offices, in particular, must take some responsibility for hyped findings. Pushed by university leaders to maximize total coverage, press officers fill their releases with claims of significant breakthroughs. Scientists then complain when their work is hyped beyond its true worth.

Researchers could also examine how other professions manage the media. Politicians are misrepresented more frequently and significantly than scientists. But they know that attacking journalists is a short-sighted strategy. Instead, they have become experts in rebutting inaccurate stories and imparting their own message.

How far should scientists go down this road? Scientists, as with almost every other profession, enjoy more public trust than politicians. Gaining a reputation for spin could damage this. Researchers should, however, learn what the media wants. Politicians understand the kind of stories that journalists are looking for. If more scientists did too, they would be better equipped to get their message across.

Some progress has already been made in this direction. Many grant-awarding bodies now promote media training for scientists. Britain's Engineering and Physical Sciences Research Council, for example, recently announced plans to include £500 (US$720) for media training in each grant it awards.

Better-funded initiatives of this sort are needed, and inaccurate reporting should not go uncorrected. But attacks on journalists often sound bitter, and have little long-term benefit. Rather than shooting the messengers, scientists should take them to one side and give them the real story.