Abstract: Among the low attitude and ground targets, rotorcrafts are the most important to be recognized in the airplane-carried millimeter wave pulse Doppler radar. Rotorcraft targets suffer from low signal to clutter rate and weak characters at millimeter waveband because the width of modulated spectrum of its radar echoes is inversely proportional to the signal wavelength. This paper first analyzes the feature of rotorcrafts in the millimeter-wave radar. And aiming at the involved rotorcraft recognition we propose some new requirements for parameters of the radar system, such as the beam dwell, pulse repeat interval, performance of clutter suppression, number of coherent integration pulses and phase noise spectrum density of the frequency synthesizer. Then, the specific target recognition platform is designed and the main structure of the software oriented at engineering application is given. Some key signal processing technologies are proposed such as spectrum pretreatment technology with noise normalization, accumulating strategy of recognition results and the ripple slice mapping of Doppler spectrum etc. Finally, the experimental data are used for feature extracting and target recognizing, and some medium results are discussed in detail. The experiment results demonstrate that the theoretical analyses in this paper are efficient and the proposed methods perform well in recognition of rotorcraft and fixed-wing aircraft targets.