This study examined genetic variation in the major histocompatibility complex(MHC) Class II B gene in turbot(Scophthalmus maximus) by virulent bacterial pathogen challenge.One hundred fry from each of six families were infected with Edwardsiella tarda by intraperitoneal injection.Family mortality ranged from 28.0% to 83.3%.Complete exon 2 and intron 1 sequences of MHC Class II B genes were amplified from five survivor and five non-survivor individuals per family using the clone-sequence method.Thirty-seven sequences from 60 individuals revealed 37 different alleles,25 of which were unique to this study.The 25 unique alleles belonged to 16 major allele types.Nine alleles were used to examine the association between alleles and resistance/susceptibility to disease.Five alleles were present in an individual,suggesting a minimum of three loci or copies of the turbot MHC Class II B gene.The rate of non-synonymous substitution(d N) was 2.30 and 1.58 times higher than synonymous substitution(d S) in the peptide-binding regions(PBR) and non-PBR in whole families,respectively,which suggested balancing selection on exon 2 of the MHC Class II B gene in turbot.One allele,Scma-DBB1*02,was significantly more prevalent in survivor stock than in non-survivor stock(P=0.001).Therefore,this allele might be associated with resistance to bacteria.A second allele,Scma-DBB1*10,was significantly more prevalent in non-survivor stock(P=0.021),and is likely associated with susceptibility to bacteria. This study examined the genetic variation in the major histocompatibility complex (MHC) Class II B gene in turbot (Scophthalmus maximus) by virulent bacterial pathogen challenge. One hundred fry from each of six families were infected with Edwardsiella tarda by intraperitoneal injection. Family mortality ranged from 28.0% to 83.3%. Complete exon 2 and intron 1 sequences of MHC Class II B genes were amplified from five survivor and five non-survivor individuals per family using the clone-sequence method. Thirty-seven sequences from 60 individuals revealed 37 different alleles 25 of which were unique to this study. 25 unique alleles were used to examine the association between alleles and resistance / susceptibility to disease. FIVE alleles were present in an individual, suggesting a minimum of three loci or copies of the turbot MHC Class II B gene. The rate of non-synonymous substitution (dN) was 2.30 and 1.58 times higher than synonymous substitution ( d S) in the peptide-binding regions (PBR) and non-PBR in whole families, respectively, which suggested balancing selection on exon 2 of the MHC Class II B gene in turbot. One allele, Scma-DBB1 * 02, was significantly more prevalent in survivor stock than in non-survivor stock (P = 0.001). Beforefore, this allele might be associated with resistance to bacteria. A second allele, Scma-DBB1 * 10, was significantly more prevalent in non-survivor stock = 0.021), and is likely associated with susceptibility to bacteria.