It is now six years since the US National Academy of Sciences recommended a new regime of community self-regulation of scientific conduct, five years since the Congress established a Commission on Research Integrity to come up with something better, and three years this November since that commission, chaired by Kenneth Ryan, a former professor of obstetrics at Harvard, issued a report calling for new structures to investigate and adjudicate allegations of scientific fraud.

Ryan's findings found some favour among university administrators, who are currently saddled with what passes for authority over the problem, but were lambasted by scientific leaders. John Dingell (Democrat, Michigan), the bête noire of the US scientific community on this and other matters, had lost his bully pulpit at the head of the Commerce Committee in the House of Representatives in 1995, and the proposals died for lack of support in Washington.

As a correspondent writes on page 823 of this issue, the resultant impasse is undesirable. Yet it is set to continue until such time as a serious incidence of scientific misconduct reignites public concern on the issue. At the rate at which significant cases are arising (see, for example, page 817), that may not take too long.

President Bill Clinton's National Science and Technology Council was asked, in the aftermath of the Ryan report, to prepare a new and universally acceptable definition of scientific misconduct. It has now been sitting on this admittedly difficult problem for two-and-a-half years. It is high time that the council published its agreed definition, and demonstrated that the US government has the courage to tackle this serious problem.