This study estimates and analyzes the structure of trade in value-added (TiVA) of world agricultural and food products. We construct a multi-regional input output (MRIO) table using the GTAP 9 DB for the analysis. World total agricultural and food trade amount in terms of value-added, bilateral trade in value-added flows of commodities, and country and product specific global value chain (GVC) participation rates are estimated. We found that the conventional trade statistics overestimate total trade volume even for agricultural and food products substantially. Our TiVA-based statistics revise the conventional agricultural bilateral trade flows significantly. Finally, our estimates found that the GVC participation rates in the agricultural and food sectors are higher for the countries with larger backward participations rather than for the countries with larger forward participations.