As a prevalent phenomenon in financial markets, the herding behaviour of institutional investors has attracted considerable research interest. There are competing hypotheses with respect to how herding behaviour is associated with information processing. On the one hand, herding weakens information collection activities when institutional investors herd in the face of information cascades (Banerjee, 1992; Bikhchandani, Hirshleifer, and Welch, 1992; Welch, 1992), relative-performance induced agency motives (Roll, 1992; Brennan, 1993; Admati and Pfleiderer, 1997), or reputation-based mimicking motives (Scharfstein and Stein, 1990; Trueman, 1994; Zweibel, 1995; Prendergast and Stole, 1996; Graham, 1999). On the other hand, fund managers simultaneously and independently move when they receive correlated information shocks (Froot, Scharfstein, and Stein, 1992; Hirshleifer, Subrahmanyam, and Titman, 1994), which speeds up information incorporation. Thus, it is an empirical question as to how mutual fund herding is associated with information processing. We use stock price crash as a consequential outcome to further examine this question. When herding fund managers become less active in collecting and processing information, their monitoring becomes less effective. As a result, corporate managers can withhold bad news more easily, increasing the likelihood of a stock price crash. However, if herding is associated with more active information collection, fund managers become more effective monitors, thus reducing information hoarding and stock price crash risk. This paper investigates the variation in the information acquisition activities of mutual funds and focuses specifically on herding behaviour. We use stock price crashes as the outcome variable to examine how varying mutual fund herding behaviours are associated with information disclosure. Mutual funds, whose portfolio holdings are subject to mandatory disclosure regulations by the U.S. Securities and Exchange Commission (SEC), provide us with a great setting to study how the herding behaviours of institutional investors affect the disclosure strategy and consequently the stock price crash risk of their holding companies. Our main results show that mutual fund herding (especially buy-herding) is positively associated with stock price crashes. This positive relationship is consistent with the information blockage hypothesis. We also find supporting evidence that mutual fund (buy) herding is associated with deteriorated disclosure quality.