THE UNITED STATES is under pressure at Geneva, where a conference on nuclear nonproliferation is going on, to suspend its underground nuclear testing. The Reagan administration, which previously rejected a Soviet call to join the Kremlin's unilaterally declared test moratorium, also has rejected the Geneva appeal. It insists that its large and worthy goal of arms reductions is not served by an end to tests, regardless of whether such an end is declared or negotiated.

In a basic sense, that's true. Tests are one thing, cuts another. As long as either side has nuclear arms, at least some minimal testing is going to be required just to keep up stockpiles. And as long -- a long time, we hope -- as either side is determined to reshape its nuclear forces in the interests of greater strategic stability, tests are also going to be necessary. Only in a world of no nuclear arms -- a dream world evoked in turn by Mikhail Gorbachev in his call for a moratorium that will lead to ultimate nuclear disarmament and by Ronald Reagan in his commitment to Star Wars -- is there no need or place for testing.

Mr. Gorbachev makes the seemingly plausible argument that tests themselves are "a kind of accelerator of the nuclear arms race." Certainly it is so that tests are essential to make new weapons; without tests, weapons development would inevitably be cramped. But other considerations -- first of all, political tension and fear -- are the real "accelerator." This is why arms control is so difficult: the politics. It is why the partial test ban treaty of 1963, leaving the superpowers free to test underground, did not curb weapons development. Much more than a mechanical shutting off of tests is entailed.

The further claim is made that a testing moratorium would sweeten the atmosphere for arms control negotiations. Would it? The year-long mutual testing moratorium that the United States and the Soviet Union agreed to in 1958 expired without bringing the negotiated comprehensive test ban it was intended to facilitate. When the Soviets, lamely citing French tests, resumed testing in 1961 with a massive series, many Americans felt betrayed by the very idea of a moratorium. True, President Kennedy declared one unilaterally in June 1963, to "help us achieve" a formal binding treaty. But almost certainly the partial test ban treaty that followed would have come anyway -- as a calming, mutually sought response to the mutual fright of the Cuban missile crisis.

So we do not think the administration should be faulted for rejecting calls to suspend or ban tests. Rather, it should be held accountable for achieving its stated negotiating goal of achieving arms reductions. Here is where solid progress must be made.