The invention relates to a biological heuristic algorithm based stealth target identification method of a cognitive radar, belongs to the technical field of cognitive radars, and aims at providing a biological heuristic algorithm based waveform optimization method which is simpler and more convenient and has a better effect compared with a present method. According to the method, a target scattering coefficient is estimated by using waveform information received by a radar receiver, maximal posterior probability estimation is used and then Kalman filtering is used for iterative estimation, theestimated target scattering coefficient is compared with a practical target scattering coefficient, a minimum mean squared error therebetween is calculated, a biological heuristic algorithm is used for solving by taking the sending waveform as a variable and the minimum mean squared error as a target function, and the sending form when the target function is minimal is obtained. The method is mainly applied to design and manufacture of the cognitive radar.