In view of the spectrum scarcity of unmanned aerial vehicles' (UAVs) communication systems, cooperative spectrum sensing (CSS) is used to identify the available spectrum for cognitive UAV networks (CUAVNs). Due to the flexible locations of flying UAVs, we first design an intraframe cooperation way to replace the traditional interframe cooperation to achieve CSS in this article. Considering certain performance requirements in CUAVNs, we perform an in-depth analysis of the relationship among performance indices of sequential probability ratio test (SPRT)-based CSS and approximate the actual detection performance and the stopping time by ignoring the excess over the boundaries and the random walk. Finally, numerical results corroborate the effectiveness and correctness of our theoretical analysis, and comparison results represent the superiority of our proposal.