This letter proposes a precise method by which to detect target motion displacement with angle migration implemented in a multi-input-multi-output (MIMO) frequency-modulated continuous-wave (FMCW) radar system. Angle migration induces phase errors when a target phase is determined on a conventional range-angle map generated by the 2-D fast Fourier transform (FFT). To mitigate these errors, the variation of the target angle is estimated first, and digital beamforming is then conducted adaptively based on the estimated target angle. This method allows extracting target motions without phase discontinuities caused by angle migration. The proposed method is verified by experiments involving both human body movement and respiratory motion tracking.