An attempt to determine how disarmament measures affect the economics of military power. The study suggests that although the broader the definition of weapon systems limited by an arms control measure, the greater is its effect, if successive expansion in scope is possible. If one agreement is expected to make successive ones easier, the expected arms control effect of a measure of narrow scope is greater than would otherwise be the case. "Narrow" arms control measures may, under the restricted conditions assumed, tend to be mutually reinforcing both in a simple sense and in that they may make secret, illicit capability disproportionately difficult. The fact that a weapon system is inefficient, and not used, does not necessarily mean that its proscription would have an expected arms control effect of zero. Some tentative implications for research are also suggested.