We study the volume–volatility relation by splitting volume into the number of trades and the average trade size at individual and institutional level, and realized volatility into its continuous and jump components. We find that the number of trades is the most important variable driving realized volatility. The number of trades by the individual investors carries more explanatory power in explaining volatility compared to the number of trades by institutional investors. The average trade size cannot be entirely disregarded, though; its importance compared to the number of trades is limited. Order imbalance plays a minimal role in driving volatility.