Although CNC (Computer Numerical Control) machinery have many technological and economic advantages in wood machining, they have serious health problems especially in case of noise pollution level (NPL). In this research, the effective parameters on NPL in CNC wood milling are investigated and optimized after modeling. For this purpose, the design of the experiment was done based on the RSM (Response Surface Method) and subsequently 27 experimental tests were performed. The results of the analysis of variance shows that the parameters of cutting speed, feed speed and depth of cut have a great impact on the noise level (respectively). In general, increasing depth of cut, feed rate and cutting speed leads to increasing NPL. Cutting speed has more effect than others. The step over does not have a significant effect on the NPL. By changing machining parameters amount of noise level changes from 97.4 to 103.8 dB (= decibel). Also in 100% of experimental tests noise level is higher than NIOSH standard value (85 dB) (NIOSH = The National Institute for Occupational Safety and Health). Therefore, it is necessary for all CNC wood workers to be entered in a hearing loss prevention program (HLPP). In the next step, based on the RSM, a quadratic regression equation was derived and optimized by genetic algorithm (GA) after validation. By choosing the optimum parameters depth of cut = 3.2mm, feed rate = 58mm/s, cutting speed = 12730 rpm and step over = 6.4 mm the level of noise pollution is reduced to 96.2 dB. This prediction tested and validated by experimental test (95.8 dB).