THREE MILE ISLAND, toxic waste, genetic engineering -- you don't have to go very far down the list of science-related public issues to know why good scientific advice is especially important in government these days. And it is this same proliferation of issues that makes the National Academy of Sciences (including the Academy of Engineering and Institute of Medicine) a potentially invaluable institution. Because of its membership and because of its stature among scientists, it can call on the voluntary services of the best of American science. Unlike even the most affluent "think tank," it does not have to worry about meeting a payroll or alienating commercial interests. Its charter (signed by President Lincoln) enables it to respond directly to the needs of government.

But this is not to say that the academy does any of these things now. The trouble is, it doesn't. In recent years the academy has tried to do too much and, in so doing, has compromised the quality of what it can do best. It now supports more than 1,000 committees, covering subjects that range from the vital to the truly trivial. Almost any request for a report or a study that comes accompanied by funds is accepted; many should not be. The huge number of committees exceeds the number of expertly qualified scientists to serve on them. And in growing so big, the academy has gotten sloppy about the two procedures that are essential to maintaining the intellectual integrity of its work.

The first of these is the choice of individuals to serve on a study.The academy duly examines detailed conflict-of-interest forms, but what really matters are intellectual biases. These are much harder to ferret out and much more important to balance or neutralize in the choice of committee members. The second and equally important procedure is the choice of individuals to review draft reports. Critical review by a panel of peers is absolutely vital to the conduct of good science. Obvious flaws in these two processes were largely responsible for the recent report by Food and Nutrition Board that has so badly damaged the academy's reputation.

The academy's president has tremendous power to shape its role, and serve for up to 12 years. The incumbent, Philip Handler, will retire next year, and the search for his successor is already under way. That choice is up to the academy's members. But considering its importance to a much larger constituency, the qualifications and views of the candidates should be the subject of a much wider debate. The workings of that club require more scrutiny than they have ever received before.