Serious scientific misconduct is a pernicious and under-reported problem that the research community has sought to police, with mixed success. There is little evidence, fortunately, that the falsification or fabrication of research data is becoming more frequent. However, there are rumblings that one type of misconduct is on the increase: plagiarism, made easier by the rapid electronic dissemination of research results (see page 258).

In many countries, the investigation of plagiarism and other misconduct remains fraught with difficulty. Universities are terrified of the adverse publicity that will accompany any finding of misconduct. And if the accused doesn't accept the outcome, they have the option of taking legal action against the accusers, the investigators or the institution in question.

In those circumstances, it is inevitable — although still unfortunate — that some senior university officials will opt for the easiest path, rather than the most righteous one. In some cases, an errant academic may be encouraged to leave an institution without full public disclosure of the events leading to the departure — a lamentable outcome that will simply serve to export misconduct, rather than to deter it.

In the United States, some defences are in place against this kind of thing. Biomedical research is overseen by the Office of Research Integrity (ORI) at the health department; research supported by the National Science Foundation, which funds most non-biomedical university research, is overseen by the Inspector General's office at the foundation. The ORI, in particular, has been active in helping universities run their own investigations properly. And if internal university probes fall short, these bodies can step in and carry out the investigation themselves.

The system established in the United States has never been quite as rigorous or effective as its architects had hoped. Yet the existence of these oversight organizations still puts the United States in a better position to police misconduct than most other nations.

In Britain, the universities themselves, through a group called Universities UK, are developing plans to establish a central office to oversee biomedical research. But the plans have already run into criticism for failing to let the body pursue its own investigations, and for the odd suggestion by the universities that the office accept sponsorship from the pharmaceutical industry (see Nature 434, 263; 2005 10.1038/434263a ).

The major French research agencies have committees for dealing with misconduct, but none of them are as visible, well staffed or effective as their US counterparts, and Italy has no such arrangement at all. Germany has three full-time ombudsmen on hand to supervise university investigations. The German system was put in place after a prominent case of data fabrication in 1997, when it emerged that dozens of studies by Friedhelm Herrmann and Marion Brach, cancer researchers at the Max Delbrück Centre for Molecular Medicine in Berlin, contained falsified data.

It shouldn't take a recurrence of something like the Herrmann and Brach case for other leading scientific nations to ensure that they have adequate measures in place to investigate misconduct, properly publicize the findings, and punish the culprits.