Plasma reactor was used to generate a high potential difference between two surfaces of concentric pyrex tubes by electrical current. The annular gap of the reactor was calculated by trial and error from the breakdown voltage equation and set at 0.45 cm. The overall objective of this research was to know the effects of the frequency, humidity, and residence time on the formation of nitrogen oxides in a plasma reactor. The primary voltage varied from 50 to 90 volts and the frequency was varied in increments of 10 Hz from 60 to 650 Hz at the primary voltage of 90. The increase in the secondary voltage was not linear but exponential at high frequencies. At a maximum concentration of about 745 ppm, the frequency and secondary voltage was 600 Hz and 4,200 volts, respectively. All tests for the effects of humidity on NO$_{x}$ production were performed at the optimal setting of 90 colts and 600 Hz frequency. Since the NO$_{x}$ production was not an one dimensional phenomenon, competing reactions were assumed to occur in the discharge chamber. The sharp peak concentration of 1,810 ppm was observed at 38% of relative humidity, The enhanced production was choked off, and the production rate rapidly dropped to 3 ppm at above 40% of relative humidity. It is assumed that the corona attacks the most vulnerable molecules in the reaction chamber before attacking other more lightly bonded molecules, possibly at humidities above 38% and the optimized 90 volt setting. Thus, there was not enough energy left after attacking all water molecules to decompose an appreciable amount of $N_2$. If nitrogen breakdown does not occur, then oxides of nitrogen are not likely to be produced.ced.