We examine the dynamics of liquidity using a comprehensive sample of U.S. stocks in the post-decimalization period. Motivated by a continuous-time inventory model, we compute a high-frequency measure of order imbalance volatility to proxy for the inventory risk faced by liquidity providers. We show that high-frequency order imbalance volatility is an important driver of liquidity and explains the often positive time-series relation between spread and volume for large stocks, which seems to run counter to most theoretical models. Furthermore, order imbalance volatility is priced in the cross-section of stock returns.