Radar noncontact detection of life signals is severely affected by factors such as environmental noise and the micro movement of detecting object, and the Doppler signal nonstationarity that caused by the interference makes the useful life signalsbreathing and heartbeat signals cannot be extracted with general signal processing method(such as digital filtering, fast Fourier transform). According to this problem, the mathematical model of radar noncontact measuring life information has been established. Using Kband, frequency 24 GHz ,power 30 mW continuous wave radar system has measured a 30 yearsold healthy men. Measurement is carried out in a quiet office environment. There is 2 m distance between radar transmitting head and measure target. Data acquisition card′s sampling frequency is 100 Hz. The data measured are carried out eight scale wavelet decomposition and reconstruction by db3 wavelet basis function. It is success that extracted the person′s breathing and heartbeat signals on the seventh lowfrequency layer and sixth highfrequency layer respectively. Through fast Fourier transform processing of these two layers′ signals, the energy frequency value of the peak area is 0.29 Hz and 1.07 Hz. The frequency difference between the traditional measuring method and the wavelet transform method is 0.04 Hz and 0.03 Hz. The results show that the wavelet transform is used to extract and separate breathing and heartbeat in radar life signals is feasible. It will provide theoretical basis and experimental data for radar noncontact measurement method applied to the clinical monitoring and disaster rescue and so on.