ProblemThis study analyzed the contribution of DRB1, DQB1, and DPB1 HLA class II alleles and resulting 3‐locus haplotypes to the overall risk of idiopathic recurrent pregnancy loss (RPL).Method of studyThe study participants included 188 Tunisian participants comprising 84 women with established diagnoses of RPL, and 104 matched multiparous females as controls. Genotyping of HLA alleles was done by PCR‐SSP.ResultsThe allelic frequencies of DPB1, DRB1, and DQB1 were consistent with Hardy‐Weinberg equilibrium among control subjects. Among class II alleles, DRB1*04:01:01 (p = .03), DRB1*08:01:01 (p = .04), and DPB1*01:01:01 (p = .04) were positively associated, whereas DPB1*04:01:01 (p = .012) and DPB1*14:01:01 (p = .006) were negatively associated with risk of RPL. The association of DRB1 and DPB1 alleles with RPL was lost after correction for multiple comparisons. The prevalence of the seven determined DQB1 alleles was not significantly different among case and control groups. Higher prevalence of DRB1*07:01:01∼ DQB1*02:01:01∼DPB1*04:01:01 (0.0417 vs. 0.000; p = .049) coupled with lower prevalence of DRB1*13:01:01∼DQB1*06:01:01∼DPB1*02:01:01 (0.0167 vs. 0.1000; p = .014) haplotypes was noted in women with RPL, thus could be considered as an RPL at‐risk and RPL‐protective haplotype, respectively.ConclusionThe study establishes weak/no contribution of class II alleles and corresponding 3‐locus haplotypes to the altered risk of RPL among Tunisian women and does not confirm previous findings on other Arab‐speaking populations of an HLA association with RPL.