Cognitive radio will drive the future of wireless/radio communication by efficient utilization of underutilized spectrum (vacant space). Spectrum sensing and primary user detection is the main challenge in cognitive radio systems. Energy based sensing is a common spectrum sensing method for cognitive radio since the information about the licensed user signal are not required for the detection process. In this paper, the performances of the energy detection technique are analyzed over Gaussian Noise Channel and different fading channels such as Rayleigh and Rician Channel based on Receiver Operating Characteristics (ROC) and probability of detection (P d ) versus Signal to Noise Ratio (SNR) curves through simulation using MATLAB. The analysis is validated with theoretical results.