The receiver operating characteristic curve is evaluated for a cognitive radio system with Laplacian noise. The characteristic functions and cumulative distribution functions at the output of an energy detector are derived for primary user ON/OFF hypotheses. Approximate closed-form expressions for the probability density functions and cumulative distribution functions at the output of an energy detector are also obtained using the Pearson approximation. With Gaussian noise, it is known that the 1-out-of-N data fusion rule provides the lowest probability of miss detection for a given probability of false alarm among K-out-of-N fusion rules. Our results show that this is not necessarily true with Laplacian noise.