Falling is one of the major health threats to the independent living of elders, and falling has received much attention in academia and industry over the past two decades. Previous fall detection methods either require specialized hardware or invade people's daily lives. These limitations make it difficult to widely deploy fall detection systems in houses. On the other hand, a long static case is also one of the health threats to the elderly; these cases indicate that something is abnormal. In this paper, we present the design and implementation of a motion detection system based on passive radio frequency identification tags. The key finding is that static, regular action, and sudden falls make an impact on the RSS and Doppler frequency values differently. Such features help us to detect the status of the elderly. A wavelet transform is used to pre-process the signal data, and a support vector machine is adopted to improve the accuracy of the fall detection. We implemented a prototype monitoring system called TagCare and conducted extensive experiments to demonstrate the accuracy and efficiency of TagCare in movement detection and fall behavior identification for the elderly.