Four greenhouse leaf inoculation methods for screening Japanese plum (Prunus salicina L. and hybrids) for resistance to Xanthomonas campestris pv. pruni (Smith) Dye were compared for repeatability, ability to differentiate among plant genotype responses, and correlations with field ratings. Clonally propagated trees were inoculated artificially in a greenhouse by immersing leaves in 2.5 × 108 cfu/ml inoculum (DIP), rubbing the adaxial side of leaves with a slurry of 2.5 × 108 cfu/ml inoculum and Carborundum powder (CARB), infiltrating leaves with 5 × 105 cfu/ml inoculum using a needle-less syringe (INFS), and infiltrating with 5 × 106 cfu/ml inoculum (INF6). No greenhouse method was superior in all assessment categories. The CARB method was most repeatable (t = 0.78) but had a low Spearman's correlation (rs = 0.29), indicating that greenhouse rankings did not correspond closely with field rankings. The INF6 method was unsuitable because it did not differentiate between plant genotypes. The DIP method appeared most suitable, having moderate repeatability (t = 0.46) for four observations per leaf and moderate Spearman's correlation with field performance (rs = 0.56). The INF5 method may be appropriate for identifying bacterial spot resistance that is associated with resistance in the leaf mesophyll.