Opportunistic spectrum access feature of cognitive radio systems is a method to improve frequency underutilization of wireless spectrum. One of the techniques for detecting the unused bands is the energy detection for which selection of the threshold defines detection performance. In this paper, the adaptive threshold method is proposed as an alternative approach to estimate the threshold as a function of first and second order statistics of recorded signals. The proposed method does not require estimation of noise variance or signal to noise ratio and aims to minimize the effects of impairments introduced by wireless channel and non-stationary noise. The simulation results indicate that adaptive threshold has low false alarm and missed detection rates that can satisfy the detection requirements of multi-channel cognitive radios for either narrow or wideband spectrum sensing, when the standard deviation coefficient is selected properly.