The most important confirmed genetic factor of susceptibility to multiple sclerosis (MS) has been identified in the HLA class II region. The hypothesis that several genes, including HLA class II, may influence the prognosis of patients with MS has been proposed. In a recent study, using low intermediate resolution typing, we found that some HLA alleles may predict disease severity as assessed by magnetic resonance imaging (MRI) measures. The aim of this study was to examine the relationship between high-resolution typing of HLA alleles and disease severity as measured by brain MRI quantitative markers of demyelinating and destructive pathology in patients with MS. In 41 MS patients (27 relapsing-remitting, 7 secondary progressive, and 7 primary progressive), we performed high-resolution typing of alleles HLA-DRB104, -DQB103, -DRB115, -DQB106, and of haplotypes -DRB104-DQB103 and -DRB115-DQB106. These alleles and haplotypes were associated with higher susceptibility to MS in a recently published case-control study conducted in the Friuli-Venezia-Giulia region, Italy. Of 41 included patients, 13 were men and 28 were women. Mean age was 43.3 (SD 11.4) years, mean disease duration 10.3 (SD 7.8) years, and mean EDSS 2.3. DNA extraction and genomic typing were obtained with the sequence-specific primers method using primer pairs that amplified the HLA alleles. All patients underwent a 1.5-T MRI examination of the brain. Disease severity was assessed by clinical measures [Expanded Disability Status Scale (EDSS)] and MRI measures. T2- and T1-lesion volumes (LVs) and brain atrophy measures [fractions of brain parenchyma (BPF), gray matter (GMF), and white matter (WMF)] were calculated. We used general linear model analysis (GML), controlled for age, disease duration, and treatment status, to compare the MRI measures according to allele and haplotype status. The following significant results were found: HLA-DRB11501 positive patients had significantly lower GMF (0.493 vs 0.526, p < 0.001), lower BPF (0.784 vs 0.815, p = 0.018), and higher T1-LV (2.8 vs 0.7ml, p = 0.036); -DQB10301 positive patients had significantly higher T2-LV (34.1 vs 0.7 ml, p = 0.041), and showed a trend for lower BPF (0.790 vs 0.846, p = 0.064); -DQB10302 positive patients had significantly lower T1-LV (2.4 vs 0.9 ml, p = 0.016); and -DQB10602 positive patients had significantly lower GMF (0.492 vs 0.521, p = 0.007) and lower BPF (0.781 vs 0.811, p = 0.023). No differences were found in the indices of MRI disease severity according to HLA haplotype associations. Both in correlation and in regression analyses, we observed significant associations between HLA-DRB11501 and lower GMF and BPF and higher T1-LV, between -DQB10301 and higher T2-LV and disease duration, between -DQB10302 and lower GMF and higher T1- and T2-LV, between -DQB10602 and lower GMF and BPF, and between -DQB10603 and higher T1-LV and EDSS. High-resolution HLA genotyping analysis revealed a robust relationship between alleles HLA-DRB11501, -DQB10301, -DQB10302, -DQB10602, and -DQB10603, and more severe damage on inflammatory and neurodegenerative MRI measures.