To develop a commercial-scale PCR-based assay for Acidovorax avenae subsp. citrulli in watermelon seed, parameters for immunomagnetic separation (IMS) were optimized. Optimal conditions for target cell recovery included 40 μg of polyclonal anti-AAC antibody per 10 immunomagnetic beads (IMBs) for coating, ca. 2.5 × 10 IMBs per sample and immunocapture for one hour at 4°C. These parameters consistently facilitated the recovery of A. avenae subsp. citrulli cells from suspensions containing ca.10 CFU/ml. Using lots contaminated with artificially infested seeds (10 A. avenae subsp. citrulli CFU/seed), IMS-PCR detected the pathogen in 25% and 87.5% of samples (n=10,000 seeds) with 0.01 and 0.1% infestation, respectively. For seedlots with similar infestation levels, the detection frequencies for the seedling grow-out assay (SGO) were 12.5% and 37.5%; however, the difference in detection frequency between SGO and IMS-PCR was not statistically significant. While liquid enrichment improved the detection sensitivity of IMS-PCR 100-fold, the difference in detection frequency between enrichment IMS-PCR and IMS-PCR was not statistically significant for seed samples contaminated with artificially and naturally infested seeds. These results suggest that IMS-PCR has great potential as an effective alternative to SGO for the detection of A. avenae subsp. citrulli in commercial watermelon seedlots.