Just how many Yugoslav tanks did U.S. and other NATO aircraft hit during the 78-day bombardment of Kosovo last spring?

That question continues to puzzle top Pentagon officials amid indications that far less damage was done to Yugoslav military equipment than NATO commanders originally claimed.

A Pentagon review of the war has been slow to produce a revised assessment. Initially due by Labor Day, the review has been extended one month to fill in lingering gaps, including a more realistic count of Yugoslav armored vehicles and artillery batteries destroyed.

Deputy Defense Secretary John J. Hamre, who is overseeing the study along with Air Force Gen. Joseph W. Ralston, vice chairman of the Joint Chiefs of Staff, has insisted on an updated battle damage assessment to satisfy questions from Congress and news organizations. The U.S. European Command is preparing a new estimate under orders from Army Gen. Wesley K. Clark, NATO's top commander and the head of U.S. forces in Europe.

But some military officials are less than enthusiastic about the effort, bristling at the attention being paid to the tally. "It isn't central to something we think will tell us how to fight the next war better," one senior Joint Staff officer said.

When the war ended three months ago, the Pentagon put the destruction at 120 tanks, 220 armored vehicles and 450 pieces of artillery. But military survey teams sent to Kosovo found significantly fewer tank hulks and other ruined equipment than expected. Clark recently told a senior Pentagon officer that analysts have verified only about 70 percent of the reported hits on armored vehicles.

In war, early damage assessments often are overstated. And in the Kosovo case, Yugoslav forces made extensive use of decoys to fool NATO pilots flying several miles above.

Allied aircraft had greater success pummeling such fixed sites as airfields, bridges, government ministries and petroleum storage facilities than mobile, easily hidden ground units in Kosovo. Besides, Air Force commanders were much less interested in going after tanks and artillery than electrical power stations, communications networks and other targets considered of greater strategic value to Yugoslav President Slobodan Milosevic.

"I don't believe Milosevic cared how many tanks he had or ever lost," a high-ranking Pentagon officer said.

Instead, this officer said, Milosevic was driven to relent by the damage NATO attacks were doing to Yugoslavia's economy and physical infrastructure. Other factors also contributed, including the acceptance of NATO's conditions by Russia, which had backed Milosevic, and the growing threat of a NATO land invasion.

LESSONS LEARNED: When the Pentagon's after-action review finally comes out, defense officials who have seen preliminary drafts say it will contain many details but few surprises or grand conclusions. Offering a preview of some of the main lessons learned, Defense Secretary William S. Cohen said last week that the air operation underscored the need for more precision munitions and intelligence-gathering systems. As NATO's first war, Cohen added, it also showed the shortcomings of coalition decisionmaking and the alliance's need to "retool its existing political machinery to be more effective for the staccato timing of a military contingency."

But Hamre, speaking to an air power symposium on Capitol Hill, cautioned against expecting any lessons to generate "major fundamental changes" in Pentagon plans, at least in the near term. He forecast only some changes "on the margin."