Recently on this page, Brent Scowcroft, John Deutch and R. James Woolsey presented arguments in favor of the small, single-warhead missile, the Midgetman ("Midgetman: Keep It on Track," April 1). They emphasize its mobility, which they rightly see as promising low vulnerability and a high degree of flexibility.

The main problem is the cost for what we would and operate for 10 years 500 missiles with 500 warheads. To put these numbers in perspective, there are about 12,000 nuclear warheads in our planned long-range force of which about 5,000 will be relatively well protected (in submarines or on alert bombers). This force costs annually about $40 billion. Therefore, per warhead, the Midgetman would cost about three times the average for the rest of the force.

Is it worth the cost? If we need 500 more warheads, and it is hard to see why we should, three additional Trident submarines would enable us to keep that number constantly at sea at about one- fifth the cost.

There are reasons for having land-based missiles. We now plan on having 950 Minuteman and 50 M-X ICBMs (with 2,450 warheads -- all vulnerable to attack) for the indefinite future in any case. The value added by 500 low-vulnerability Midgetmen to this total might be worth it if the cost is sharply reduced. As a recent report released by Rep. Les Aspin (D-Wis.) shows, this $50 billion bill could be reduced by about $20 billion if three warheads were put on each missile.

Why is it better to have one than two or three warheads per missile? One argument is that it helps arms control. The Midgetman has been constrained by Congress to have only one warhead in order to limit its offensive power against Soviet missile silos and to encourage the Soviets to emulate our restraint on avoiding multiple warheads. As Scowcroft et al. point out, this is an unconvincing argument.

The other argument for one warhead is that a missile with several warheads would be too heavy to be mobile. The proposed 39,000-pound one clearly would be compatible with good mobility and Scrowcroft et al. assert that a 75,000-pound one would not. I doubt that anyone knows at this point. The Air Force should find out.

The estimated weight of Midgetman with several warheads is driven up strongly by the assumption that it needs big warheads in order to be able to destroy Soviet hard targets. Restricting the Midgetman's payload to only one warhead (to limit its ability to destroy Soviet silos) while making that a big one (so that it can destroy silos) may have congressional logrolling logic, but the result is weird. (One can see why Sen. Pete Wilson (R- Calif.) has said the Midgetman should be named the "Congressman.")

Why a big warhead? This is to maximize our prompt, hard-target, retaliatory, kill capability, with land-based ICBMs. On this view, our thousands of increasingly accurate submarine missiles -- with a much-improved, hard-target kill capacity -- won't do because communication with submarines isn't reliably "prompt" enough to respond to a sudden Soviet attack. This requirement for "prompt retaliation" inexorably brings to mind the image of our launching thousands of warheads at empty Soviet silos.

They say that changing the missile design would cause a two-year delay. If that's true, what's the hurry? Although one can conceive of a greater Soviet attack potential against our silo-based ICBMs and bombers and also new threats against our submarines, these potential dangers do not seem great enough to make it urgent to get 500 cost-ineffective Midgetman warheads in the field to join the 12,000 others.

We have had two main problems in our prolonged attempt to modernize our ICBM force. One is finding a secure basing system, and the Midgetman mobile launcher looks ood on this score. The other is our preoccupation with the possibility of a huge bolt-from-the-blue Soviet attack. Of course, we must take precautions -- and we will remain vigilant -- against such a possibility; but we should be more concerned than we have been about a less obviously suicidal Soviet move: the limited use of nuclear weapons against the forces of our allies or even our own.

In short, ICBMs usefully complicate the Soviet attack problem and the mobile launcher concept looks like a reasonable solution to a long-vexing problem. But we should insist that the missile be cost-effective; whatever the maximum size permitting good mobility is, the missile finally designed should accommodate several warheads of whatever size will fit.

With defense dollars in shorter supply, we need all we can get for other pressing purposes such as improving command and control of our forces, buying modern nonnuclear weapons, and having more security assistance. Thse larger trade-offs should be examined before we press ahead with a highly expensive system per warhead.