IN THE END the National Institutes of Health got it about right. This year, NIH proposed tightening its ethics guidelines to such an extent that the rules would have unnecessarily penalized employees and ultimately harmed the ability of the institutes to attract and retain top scientists. In issuing the final version of the rules Thursday, the agency backed away from its most draconian proposals while putting in place needed restrictions on outside consulting.

Hardly anyone disputes the charge that the previous rules were overly permissive and loosely enforced. Of particular concern were lucrative consulting deals, some of them ethically questionable, between government scientists and private industry. A report issued earlier this month by the Health and Human Services inspector general found that NIH ethics officials routinely approved senior scientists' moonlighting requests without having enough information to judge whether there was any conflict of interest.

Given these difficulties, it made sense for NIH to simply prohibit outside consulting work. What didn't make sense, and what we're glad to see NIH retreat on, was a sweeping rule that barred all NIH employees -- and their spouses and minor children -- from owning more than $15,000 worth of stock in a wide array of affected companies, such as drug makers, biotechnology firms and manufacturers of medical devices. In addition, at least one-third of the NIH workforce under the proposed rule would have been prohibited from having any stock holdings whatsoever in such firms.

Instead of this overbroad and eventually self-defeating approach, the new ethics rules impose stock restrictions on only the most senior employees -- about 200 in all -- who can't own more than $15,000 of stock in the covered companies. Other employees, the new rules say, "may be required to divest if, after review, a potential conflict resulting from their holdings or those of their spouses and minor children would impede their ability to do their government job."

That case-by-case approach is appropriate, so long as it is coupled with an effective system for reviewing potential conflicts and keeping NIH out of the ethical soup in which it's been stewing for the last few years.