Like crime, scientific misconduct ultimately doesn't pay. For the two highprofile cases of data fabrication in recent years — Jan Hendrik Schön of Bell Laboratories and Woo Suk Hwang of Seoul National University — the price was a ruined reputation and dismissal. Others found guilty of similar misdeeds have found themselves barred from doing research or from receiving federal funds (see Nature Med. 12, 490–494; May 2006).

But what of the innocent parties caught up in the crossfire? Postdocs and graduate students working under a principal investigator who is found guilty of misconduct can find themselves tainted by association. At the very least, they will face setbacks in their careers — what use to any author is a retracted paper? And there is little point in asking your adviser for a letter of reference if he or she has been discredited.

People working under an adviser whom they know to be engaged in misconduct face a difficult choice. If they do nothing, then they stand a chance of completing their studies and moving on to a different lab — but if their adviser is eventually found out, they risk being judged as guilty by association. If, on the other hand, they choose to blow the whistle, life can become very uncomfortable. In many instances, young scientists who report incidences of misconduct by their seniors can face as much scrutiny as the people they have accused (see Nature 441, 122–123; 200610.1038/nj7089-122a).

To resolve these issues, and to minimize any risk of fraudulent behaviour, we need a new system that protects whistleblowers. Such a system would give universities, funding agencies and journals clear roles and guidelines for policing fraud. It would also, by necessity, shift the burden of proof from the accuser on to the accused.

If trainees find themselves in the unfortunate position of having to point out questionable research practices, they should be able to do so without worrying that they could damage their career. This is especially true when you consider that not blowing the whistle could have even worse long-term effects.