Spectrum sensing is one of the major elements of cognitive radio applications. A single secondary user (SU) usually cannot provide robust sensing capability due to channel effects. In order to overcome this problem, cooperative spectrum sensing techniques have been proposed. Because conventional cooperative spectrum sensing schemes assumed that all sensing nodes have the same sensing reliability, the SUs were assigned identical false-alarm probability even though some of them suffer deep fading, which may degrade their detection performance. In this study, we propose an adaptive credibility-based cooperative spectrum sensing technique, which evaluates the sensing reliability of SUs based on previous sensing performance, and adjusts the probability of false alarm of each SU individually. Assigning a higher false alarm probability for a more reliable SU can result in better detection performance. To verify the effectiveness of the proposed cooperative spectrum sensing technique, both the false alarm probability and the detection probability are simulated under Rayleigh fading channels, and the results show that the proposed approach outperforms any conventional detection method in terms of detection performance.