This paper investigates order-based manipulation and its effects on investor behavior and market efficiency. Using a unique data set from the Chinese stock market, we show that (1) order-based manipulation affects market liquidity and trading behavior, (2) the manipulator pretends to be informed or expects to be seen as informed by choosing a "right" time to implement the manipulation, and (3) the manipulation rapidly changes investor reaction to the market order/depth imbalance in the short run, and the effect gradually drops during the postmanipulation period. Our results are robust to alternative measures and offer clear implications for both theory and policy.KEY WORDS: depth imbalance, order-based manipulation, order imbalance, spoofing order.The traditional definition of spoofing involves the display and prompt withdrawal of a large quote or order with the intention of triggering a market movement from which a market participant may benefit by trading on the contra-side (i.e., buying at an unduly low price created as a result of a flash quote to sell, and vice versa). Recently, the definition of spoofing has been broadened to include any placement of an order or quote to be displayed for the purpose of gaining improper market advantage.Manipulation in stock markets has fascinated practitioners and academics for a long time, probably for as long as stock markets have existed. Identifying manipulation is also important for regulators, since manipulation interferes with the free forces of supply and demand and undermines the integrity of prices, investor confidence, and market justice. However, due to a lack of empirical data, most academic papers focus on theoretical models, and there are rarely studies on real-world manipulation (e.g., Aggarwal