Abstract We compared two bycatch reduction device (BRD) models that were designed to maintain the catch of commercially harvestable blue crabs Callinectes sapidus while excluding diamondback terrapins Malaclemys terrapin. The Virginia BRD (VA BRD) model was a thick-plastic, 5.1- × 15.3-cm, rectangular red frame. The South Carolina BRD (SC BRD) model was a thin-plastic, 5.1–6.4- × 7.3-cm, curved rectangular red frame. Baited crab traps were fished in groups of three (no BRD, VA BRD, and SC BRD) within tidal creeks in Virginia and South Carolina. In Virginia, legal-sized blue crabs from traps with BRDs were 2 mm smaller than those from traps without BRDs. In South Carolina, significantly fewer and smaller crabs were captured in traps with SC BRDs relative to traps with VA BRDs or traps without BRDs. Shorter soak times in South Carolina may have reduced traffic flow of crabs into traps fitted with BRDs, an effect that was overcome by longer soak times in Virginia. Of the 29 diamondback terrapins captured during the Virginia study, 23 were captured in traps without BRDs, 3 were captured in traps with VA BRDs, and 3 were caught in traps with SC BRDs. The BRDs are needed for the blue crab fishery to decrease diamondback terrapin mortality in hot spots where crabbing and terrapin habitat overlap.