This paper presents the results of experiments in which various parameters were varied systematically in an attempt to understand how the reactor design affects the energy efficiency for plasma processing of NOX. These parameters include the packing material, electrode diameter, and voltage frequency. It is shown that the applied voltage is not the relevant parameter when comparing the performance of different plasma reactors. The important control parameter is the input energy density. In accordance with the observations reported by Penetrante et al. [Applied Physics Letters 68, 3719-3721 (1996)], we have found that reactor design has little influence on the basic energy consumption of the plasma. Consequently, different reactor designs should yield basically the same plasma chemistry if the experiments are performed under identical gas composition and temperature conditions.