Technical Abstract: Rapid, reliable, sensitive, and high-throughput methods are essential for food defense, to detect foodborne contaminants and to facilitate remediation and recovery from potential toxin-related incidents. Ricin is a protein toxin that belongs to the large family of dichain ribosome-inactivating proteins and has been used for intentional contamination of foods in the past. In this study, we developed and compared procedures for quantification of ricin in food matrices using immuno-PCR (IPCR). The direct adsorption of ricin onto the wells of a microtitration plate was compared to indirect immobilization via a capture antibody (sandwich IPCR). The latter procedure provided much greater sensitivity. We also compared a protocol with the immunoassay and PCR conducted in a single plate to a two-step procedure in which the PCR was conducted in a second plate, following release and transfer of the DNA marker. The two-step procedure proved about 1000-fold more sensitive for ricin detection, and this format was used to detect ricin in spiked samples of ground beef, liquid chicken egg, and milk, and the results were compared to those obtained using a sandwich ELISA format. The IPCR had a limit of detection (LOD) of 10 pg/mL in liquid chicken egg and milk samples, and 100 pg/mL in ground beef extracts. Comparable ELISA results were in the 1 to 10 ng/mL range. Thus, IPCR affords 10-fold greater sensitivity in the ground beef matrix, 100-fold in the milk and 1000-fold in the egg matrix, compared to ELISA. Further optimization of the sandwich IPCR was performed by comparing various antibody combinations: homologous polyclonal detection and 3 heterologous mAb/pAb and mAb/mAb pairs. Among the four formats investigated, the pAb/pAb combination yielded the lowest LOD (10 fg/mL).