IT'S time to take another look at scientific misconduct. Fraud and fakery have tarnished science's image. But in its zeal to police federally funded research and mollify congressional critics, the United States government has itself become a source of misconduct.Accused criminals can confront witnesses and challenge evidence. Accused scientists enjoy no such "due process" unless they are actually in a court of law facing charges of legally defined fraud. In most misconduct cases, they are subject to a "star chamber" administrative process whose rules explicitly deny them the right to confront witnesses or examine in detail the evidence against them. Furthermore, the "crime" of scientific misconduct is ill-defined. Guidelines issued in 1986 specified plagiarism, fakery, and false representation. In 1989, that definition was stretched to include the fussy category of "other practices that seriously deviate from those that are commonly accepted within the scientific community for proposing, conducting, or reporting research." Any error or sloppy work an investigator considers deviant thus becomes suspect. Present practice would seem to deny ordinary civil rights. However, a federal judge in the western district of Wisconsin ruled last December that such rights are not involved because research grants "are not made for [a scientist's] benefit, but for the benefit of the general public." The ruling added that a scientist's "ability to serve on committees, to review journal articles, to participate in peer review, to conduct his research without supervision or special reporting obligations is not a right that is protected by law." Yet the imposition of sanctions - including loss of funding - can be as onerous for a creative scientist as any penalties imposed by a court of law. Fairness demands credible due process for the accused. It is the practices of the Office of Scientific Integrity (OSI) of the National Institutes of Health (NIH) that are specifically at issue. OSI has taken the lead in federal misconduct action. The new NIH director, Bernadine Healy, has called OSI procedures sloppy and unfair. She wants to reform them. This has drawn fire from fraud-hunting Rep. John Dingell (D) of Michigan. During hearings last month, he accused Dr. Healy of trying to dismantle OSI. This has put her reform attempt on temporary hold. Mr. Dingell and his oversight and investigations subcommittee will likely continue the hearings this fall. They should move on to the larger issue of due process. Misconduct must be clearly defined. Intentional fraud, plagiarism, and fakery warrant vigorous prosecution. But the category of deviant behavior invites witch hunting. Also, accused scientists need guaranteed rights to confront witnesses and examine evidence. Defenders of OSI procedures argue that this would discourage whistle-blowers. But that is no reason to deny justice.