Abstract This study evaluated the impact of off-center diagonal (OCD) profile depth pairing between the treatment planning system (TPS) and the electronic portal imaging device (EPID) on gamma pass rates in portal dosimetry. In clinical workflows, OCD profiles are used in the TPS to generate predicted images via the portal dosimetry image prediction (PDIP) algorithm and in the EPID system to correct measured fluence. The consistency of these settings may influence verification accuracy. Portal images were acquired using a TrueBeam linear accelerator with an aS1200 EPID for four photon energies: 6X, 10X, 6 flattening filter-free (FFF) and 10FFF. Five OCD profiles (reference depth, 5, 10, 20 and 30 cm) were configured in both the PDIP model and EPID system. For each energy, a total of 175 plan–measurement combinations were evaluated, derived from five PDIP OCD depths combined with five EPID OCD depths across seven field sizes. Field sizes ranged from 5 × 5 to 30 × 30 cm2. Gamma analysis used 3%/3 mm criteria with a 10% dose threshold. A two-way analysis of variance assessed the effects of TPS and EPID OCD depths and their interaction. For 6X and 10X beams, pass rates varied with configuration, showing better agreement when depths were matched or EPID was deeper. In contrast, 6FFF and 10FFF beams maintained high pass rates with minimal variation. These findings indicate that OCD depth pairing influences portal dosimetry performance, particularly for flattened beams, underscoring the importance of depth-aware configuration in QA protocols.