With an LFMCW automotive radar operating at its center frequency of 77GHz, the sequential estimation of frequency and amplitude for vital signs, namely, respiration and heartbeat, are considered. The radar response of vital signs is described and analyzed. With extraction of the phase history, extended Kalman filter and particle filter are simulated. Extended Kalman filter with a certain number of samples per iteration performs well for sequential estimating the respiratory frequency and amplitude in different scenarios. This sequential estimator is then verified by experimental data.