Common bacterial blight (CBB) is one of the most important diseases of common bean (Phaseolus vulgaris L.) and is caused by Xanthomonas campestris pv. phaseoli and X. campestris pv. phaseoli var. fuscans.  Backcross and pedigree methods have commonly been used to breed for CBB resistance, whereas recurrent selection (RS) has not been evaluated for improving CBB resistance.  The objectives of this study were to 1) assess the effectiveness and number of RS cycles required to pyramid high levels of CBB resistance and 2) determine the effect of intermating genotypes with differing levels of CBB resistance for RS.  A double-cross F₁ of Wilkinson 2 / DRK 2 // DRK 1 / VAX 3 of 711 plants was used as the base population (C₀S₀) in this study. Selected resistant by resistant (R x R), resistant by intermediate (R x I), and intermediate by intermediate (I x I) plants were intermated to generate three subpopulations within each RS cycle.  Three subpopulations for each of C₀S₂, C₁S₂ and C₂S₂ were evaluated for CBB reaction in the greenhouse using a randomized complete block design with four replicates in 2006 and 2007.  The CBB resistance improved with each RS cycle in all three subpopulations.  The C₂S₂ had the highest mean CBB resistance, although mean CBB scores were similar between the R x R and R x I subpopulations.  Moreover, the greatest number of resistant recombinants were produced in the C₂S₂ R x R subpopulation, followed by the C₂S₂ R x I and I x I subpopulations.  At least one cycle of recurrent selection was needed to produce a family with a mean CBB-resistant score. Additionally, it was possible to pyramid a high level of CBB resistance using RS among the C₀S₀ recombinants with an intermediate CBB reaction.