Arms makers are reaping billions of dollars more than they ought to because the Defense Department too often accepts their word for what it costs to build a weapon, the Pentagon's inspector general has concluded.

On some of the military's largest weapons programs, ranging from jet fighters to missiles and satellites, government negotiators consistently failed to challenge contractor cost estimates vigorously, he also concluded.

Examining a sample of 17 major programs, Inspector General Joseph H. Sherick found that the Pentagon could have saved $7 billion during the next five years by consistent use of "should-cost" accounting, a method of independently estimating what a weapon ought to cost. The audit said the military's buyers routinely neglect to conduct "should-cost" estimates, which are required by Defense Department regulation.

"These problems have led to lost opportunities to achieve lower contract prices," the report concluded. "On the average, contracting officers negotiated almost twice as successfully when should-cost analyses were used to prepare for negotiations."

The Sept. 26 report, signed by John W. Melchner, assistant inspector general for auditing, was made available by the Pentagon upon request this week. Defense Department auditors limited their scrutiny to whether contractors were spending all the money they said they were spending in building weapons, but did not investigate whether they should be spending so much, according to the report.

Air Force and Navy officials disagreed with some of its conclusions and urged that they not be made public. James E. Williams Jr., deputy assistant Air Force secretary, questioned the $7 billion estimate of potential savings and wrote in a memo, "Publication of such a conclusion could lead to embarrassment of the department."

But James P. Wade Jr., assistant secretary of defense for acquisition, accepted the report without dissent and said he would move to adopt its recommendations. Wade said that Deputy Secretary William H. Taft IV has initiated a test program to apply "should-cost" techniques as part of an effort to reduce contractor overhead costs.

Army Assistant Secretary J.R. Sculley also concurred with the report, "including the potential savings."

The audit appears to confirm a contention of A. Ernest Fitzgerald, an Air Force official and in-house critic who has long said that Pentagon auditors measure contractor proposals only against past costs, which may have been unreasonable, and not against independent estimates of efficient production costs. Negotiators with "should-cost" analyses in hand were able to reduce contractor proposals by 15 percent, compared with about 7 percent otherwise, the report found.

"Should-cost" accounting differs from normal cost analysis, according to federal regulations, "in its depth, in the fact that it is conducted at the contractor's plant and in the extent to which the government identifies and challenges inefficiencies."

Since 1976, it has been a required part of Pentagon contract negotiations, but the Defense Department "gave too much latitude" to each service in determining whether to follow the rule, the report says. As a result, only the Army consistently uses "should-cost" techniques.

The report found that arms makers with some of the military's largest sole-source contracts have never been subjected to "should-cost" scrutiny. These included General Dynamics' F16 fighter, RCA's military weather satellite, McDonnell Douglas' F18 fighter and attack plane, Lockheed's P3 surveillance plane and Texas Instruments' HARM missile.

The Navy objected strenuously to the inspector general's report, saying that "should-cost" analysis is only one way to reduce costs and that each service should be given flexibility to decide when to use it.

"Expanded use of competition and fixed-price contracts are the best tools we have to get reasonable prices and to limit the government cost risk," Navy Assistant Secretary Everett Pyatt wrote.

Pyatt also disputed the $7 billion estimate and said the report would "run the risk of providing specious but politically effective reasons for arbitrary cuts by the opposition and generating unrealistic expectations among our friends."