A confidential evaluation by the Pentagon warns that plans to cut the Navy's fleet of attack submarines--already down sharply to 56 from 96 a decade ago--would jeopardize critical operations and forecasts a need for 68 boats by 2015, defense officials said.

The 40-page report by the Pentagon's Joint Staff reflects the judgment not just of the Navy, but also of the military's regional commanders and senior aides to Gen. Henry H. Shelton, chairman of the Joint Chiefs of Staff. As such, it poses a fresh budgetary challenge to the Clinton administration, which is already hard-pressed to finance a national antimissile system, new generations of fighter aircraft, military pay raises and other initiatives that military leaders have pushed.

The study offers no cost assessment, but a rough calculation using Navy figures indicates the price of building up the fleet to 68 attack subs would be at least $14 billion.

Navy officials welcomed the report, regarding it as helpful in refuting suspicions that they have been inventing missions to rationalize a large fleet of attack submarines in addition to America's 18 ballistic missile subs, which are part of the nation's nuclear deterrent. The heightened demand for attack submarines, the study says, comes from U.S. intelligence officials and four-star regional military commanders who favor using the boats for high-tech spy missions and the insertion of commando teams.

"We're not a self-licking ice cream cone," said Rear Adm. Malcolm Fages, director of submarine warfare.

Just 2 1/2 years ago, during the Pentagon's last major review of defense programs, top military authorities decided to shrink the number of attack subs to 50 by 2003. But Navy officials now say there was little strategic analysis behind that decision.

In raising the preferred bar to 68, the Joint Staff report also concludes that the Pentagon could go as low as 55 attack submarines before being unable "to meet urgent critical demands." That number is one fewer than the current size of the fleet.

The Navy has struggled to dispel the notion that attack submarines--whose Cold War purpose was to shadow Soviet nuclear subs and, in the event of war, sink them before they could fire their nuclear missiles--are useless relics of an earlier era.

While acknowledging a much-diminished need to follow Russian submarines, Navy authorities say their underwater fleet is in greater demand than ever for intelligence gathering against potential Third World adversaries. These highly classified assignments include tapping into undersea fiber-optic cables, intercepting communications signals, monitoring weapon tests and spying on military exercises.

In wartime, Navy officials add, attack submarines can sneak into foreign waters and launch cruise missiles, detect undersea mines or drop special forces on enemy beaches. Last spring, for example, U.S. and British submarines fired 25 percent of the Tomahawk missiles used in NATO's air war against Yugoslavia.

But some naval experts question the need for enlarging the fleet, expressing skepticism that intelligence missions could have expanded so substantially since the Cold War, or that submarines offer the only way to carry them out.

"The intelligence stuff is difficult to assess because it's secret," said Ivan Eland, director of defense policy studies at the Cato Institute here. "But all the rogue states we say we're now worried about--Iran, Iraq, Libya and so on--existed during the Cold War, and we were doing a lot of intelligence gathering then. I can't believe a lot of new requirements have suddenly developed."

By Eland's calculations, which assume fewer intelligence missions and a halt to chasing Russian boats, the Pentagon could make do with 25 submarines.

The new Joint Staff study comes at a critical moment for the Clinton administration, which is in the final weeks of drafting its fiscal 2001 budget request. All the military service chiefs have complained of being short tens of billions of dollars for necessary projects, despite President Clinton's pledge last January of $112 billion in additional defense spending over the next six years and a congressional decision to add still more billions in fiscal 2000.

A high-ranking Pentagon general held out little hope of building or refitting enough submarines to boost the fleet to 68. But the Navy appears determined to hold the line at 56 boats and, possibly, to gain a few.

It already has embarked on a $64 billion program to build 30 Virginia-class submarines over the next two decades to replace the retiring Los Angeles class. The new model is smaller and slower than the huge--and hugely expensive--Seawolf, whose production was halted at just three boats after the end of the Cold War. But the Virginia-class subs, which have torpedo rooms that can be reconfigured for specialized missions such as deploying Navy SEAL commandos, still are pricey at $1.65 billion apiece in 1995 dollars.

Fages contended that the price of the new boat is not much more than what it would cost to reproduce a Los Angeles-class model today. While an attack submarine costs roughly twice as much as a new destroyer to produce, he added, the long-term or "life cycle" costs are about the same, because operating a submarine is cheaper. The admiral also noted that the price of the Virginia class is driven up by its low rate of production, planned at one or two a year.

To shrink the fleet, the Navy has been taking perfectly good Los Angeles-class submarines and decommissioning them years ahead of schedule. Intent now on reversing course, the Navy proposes to extend the life of eight Los Angeles-class submarines slated for early retirement and refuel their nuclear cores. Also under consideration is conversion of four of the nation's 18 nuclear-armed Trident submarines to conventional attack boats. But neither of these moves would be cheap.

The refueling would cost about $200 million per boat, including money saved from decommissioning. And converting the four Tridents would cost at least $2.4 billion and possibly twice as much, Navy officials said. The Trident plan also poses a nettlesome arms control problem: Even after conversion to attack subs, their missile launchers still might have to be counted as nuclear weapons under the START I strategic arms reduction agreement.

The Joint Staff study was requested by Deputy Defense Secretary John J. Hamre 21 months ago. Approved last month by Shelton, it remains subject to final review by the Pentagon's civilian leadership. But senior military officials said there was little dispute over the study's conclusions, which remain focused narrowly on how many submarines are likely to be needed between 2015 and 2025, not how to fund them.

Asked why the Navy doesn't design a cheaper underwater vessel devoted solely to intelligence gathering, Fages said the reduced size of the fleet and uncertainty about future threats require all submarines to be able to fight as well as spy.

Asked whether the Navy could make do with fewer submarines by increasing the time each spends deployed--now limited to one six-month stint every two years--the admiral cited various constraints. One is Navy personnel policy, which limits sea tours for sailors to six months at a stretch. Another is the life of the nuclear core; greater submarine use would result in faster fuel consumption and shorten the 33-year life span of the boats.

Still, for all their concern about the dwindling submarine fleet, Navy leaders appear reluctant to pay for more boats at the expense of cutting back on high-priority plans for more surface ships and new carrier-based aircraft. In the absence of new money from Congress and the White House, the Navy would seem to have little choice but to accept a smaller submarine force.

Even some defense analysts sympathetic to the Navy's arguments think the service may be wasting time expecting a reversal in course and would do better to start figuring out how to cope with shrinkage to 50 boats or fewer.

"I think the submarine community, because of where the defense debate in this country is going, is in a heap of trouble," said Ron O'Rourke, a naval expert with the Congressional Research Service. "I don't think they are ever going to shake the stereotype that they are part of the old architecture--at least the submarines we have on the books today."

CAPTION: DEEP REDUCTIONS (This chart was not available)