In today's competitive Industrial environment, it is very essential to achieve maximum availability of plant and machines. The Industrial equipments are subjected to regress operating conditions and therefore the possibilities of break down from various inaccuracies and problems are quite significant. Therefore, a dynamic predictive maintenance based on machinery problem diagnostics is a must in ensuring trouble free operation. Machinery vibrations are manifestation of the health condition of operating machines. Hence the accurate measurement and correct interpretation of vibrations can help in precisely diagnosing machinery problems during operation. The aim of this is to investigate how the acoustic signals can be used for state monitoring of the operating machines. The Fast Fourier Transform method is used to analyze the characteristics of the machine, based upon which the machinery problems during operation can be diagnosed precisely. The Fast Fourier Transform of and the acoustic signals are Computed. On time state monitoring in the running condition of the machine is performed, which eliminates the need for a separate day for maintenance when the machines are shut down for monitoring the machine health condition. Hence a reliable process for the state monitoring of the machine health is put forward. If the vibration amplitudes are high, the health of the machine is not good. Similarly the noise produced by the machines during the operation is also the indication of condition of the machine. Therefore amplitudes of these signals can also be used for the condition monitoring. The acoustic diagnosis of machine state is as efficient as vibration monitoring and cost effective.