The objective was to minimize the errors identified with measurement of density and moisture content of soils using nuclear gauges. To provide a reasonable basis for optimization, all identified errors were combined to yield a single criterion, the quality factor, which can be used to evaluate the over-all performance of a nuclear density gauge. Research aimed at minimizing moisture content measurement errors involved using the Monte Carlo or random walk method to simulate gauge response, checking the results of the simulation against experimental studies, and attempting to generalize the Monte Carlo results.