Human detection and identification are topics of recent interest because of the increased concerns regarding security and surveillance, and therefore has attracted the attention of many researchers. Detection of moving human can be applied in search-and-rescue operation, law enforcement, border patrol, and smart automobiles. To detect human beings, various sensor technologies have been developed. They include the use of computer vision, seismic sensor, infrared detectors, and radar systems. Among them, radars offer a number of unique advantages compared with the other technologies. Hence, in this thesis, human detection using radar systems has been investigated. In order to detect human beings, given that different radar targets have different distribution patterns, we first need to obtain information about how to model human echo. Since each of the various components of the moving human body has different radar cross sections, velocities and directions of motion, the echo signal will have different amplitudes and Doppler shifts, which should be considered in the modeling of the echo signal. In this thesis, explanations have been given about human detection based on the linear and nonlinear phase approximation, as well as the classification based on the classification algorithms, and finally, human detection from the viewpoint of the detection theory is considered for various assumptions. The results have been analyzed and compared using numerical simulations in MATLAB. Keywords: Human detection, Radar systems, Detection theory