The Mark 21 nuclear warhead has become the latest of several key parts of the MX system that apparently will have to be rushed to meet the December, 1986, initial deployment date chosen by President Reagan for the intercontinental ballistic missile, according to congressional and Department of Defense sources.

"The MX has faced some unique delays," a key Pentagon official said last week, "and we have had to compress schedules, giving us less ability to make changes."

A high-level meeting was held at the Pentagon Friday to discuss how testing, production and construction problems involving the missile, its warhead and the key facilities involved in its basing could be handled in order for the first 10 MXs to be in their silos at Warren Air Force Base, Wyo., and declared operational by the end of 1986.

"The MX is a matter of priorities, and we have to define what else might have to be given up to get it done," a key Pentagon official said Friday.

Two months ago, Defense Secretary Caspar W. Weinberger was told in a memo by Dr. Richard L. Wagner, assistant to Weinberger for atomic energy, that it was "unlikely" that the Department of Energy could have the planned new MX warhead ready "earlier than the spring or summer of 1987," according to informed sources.

This month, the Air Force ballistic missile office, which is in charge of the MX program, reported to its superiors at the systems command that it did not believe that the new warhead would be ready for planned MX flight tests. The office said it wanted to study arming the first MX missiles with the existing Mark 12A warhead rather than wait for the new one, sources said.

One Senate critic of the MX said last week that the administration is pushing the 1986 date because "any delay will push it back to near the Trident II," the submarine-launched ballistic missile that is on the drawing boards. That missile, expected to be operational in 1989, is expected to perform the same tasks as the MX, but would not have the vulnerability problems of the land-based system.

On Friday, Henry E. Catto Jr., assistant secretary of defense for public affairs, said Reagan is holding to the 1986 date because "the sooner we get the MX, the better deterrence is going to be."

"The Soviets have 600 MX-type missiles already in place," Catto said, "and we want to get ours in place as quickly as possible."

Another top Pentagon official, who asked to remain anonymous, gave another reason for pushing the 1986 date: "We manage the program by keeping the pressure on. Slips in deployment become self-fulfilling. And month delays extend into years."

The situation involving the MX warhead illustrates how problems have developed and the impact they have.

The Carter administration chose the Mark 12A warhead for the MX because it was in production. It packs an explosive punch of 340 kilotons (the equivalent of 340,000 tons of TNT, or almost 30 times the force of the 12.5-kiloton bomb used at Hiroshima in 1945), and three of them were being placed atop 300 Minuteman III missiles in a modernization program started by the Nixon administration.

Under Reagan, however, the Pentagon decided on a more powerful warhead for the MX. It will have the explosive power of nearly 500 kilotons and a more accurate guidance system than the Mark 12A. Because it is new, however, it must undergo flight tests and underground explosive tests.

For the past two years, while the administration has been having problems selecting a basing system for the MX, Congress has been refusing to provide the money needed to begin constructing special production facilities for the MX. These facilities are needed for two reasons, according to sources in the departments of Defense and Energy, which builds nuclear warheads.

First, each nuclear bomb, shell and missile requires special equipment for its manufacture. Such equipment has to be funded, designed and purchased years ahead so that it will be ready when production begins.

Second, the nation's nuclear weapons production facilities are being stretched to their limit, with the greatest number of weapons being turned out in more than 20 years. These include the 8-inch neutron artillery shell, the air-launched cruise missile, the Trident I submarine-launched missile, the B61 and B83 bombs and soon the ground-launched cruise missile and Pershing II missile.

In the fiscal 1983 Energy budget for defense programs, the administration asked for $60 million in advance funding for the MX. Last December, Congress froze the funds along with other MX development money.

Although that freeze was lifted in May when Congress approved putting the MX in Minuteman silos, a Pentagon official said Friday that the fiscal 1983 funds for the MX warhead have yet to be released.

Meanwhile, the House Appropriations Committee has cut an additional $10 million from the $30 million requested for fiscal 1984, saying the action was "consistent with delays in the decision on this system."

In the April memo to Weinberger, Wagner wrote that even with prompt congressional support with funding and cutting into other nuclear programs, the Energy Department would have difficulty getting the new MX warhead ready for the 1986 deadline.

On Friday, a Pentagon official involved in the MX program said that some other nuclear weapons program might have to be delayed to get the MX warhead built on time.

Air Force officials, however, have a different concern. They are worried that the final design for the Mark 21 warhead will not be ready early enough for them to use a precise mockup in their flight tests that are designed to test the accuracy of the 10 reentry vehicles the MX will carry.