Radio Frequency IDentification (RFID) technology plays an important role in deterring counterfeit products. There are, however, many unsolved challenges in designing an effective and efficient RFID system for anti-counterfeiting purpose. The first challenge lies in designing a system which is suitable for standard passive RFID tags, especially EPC C1G2 tags, which are widely used in real-world supply chains. These tags cannot afford strong security primitives due to very limited storage and computational capability. The second challenge lies in designing a system which can efficiently handle a increasing number of RFID products. Existing approaches for clone tag detection suffer from performance bottlenecks because a centralized detection server is required in these approaches to record and analyze the traces of all individual products. Thirdly, a practical approach to clone tag detection should work efficiently in real-world RFID systems, where products are moved and processed in batches. To address these challenges, we propose a Batch Clone Detection (BCD) scheme, which is simple, efficient, and practical. It requires only a few bits of storage in RFID tags and no computation on RFID tags, which makes it suitable for standard EPC C1G2 tags. In BCD scheme, the clone tag detection is performed at a batch level, which significantly reduces the storage and computational overheads on the server side.