The objective of the study was to examine the relationship between HLA genotypes and disease severity as measured by brain MRI quantitative markers of demyelinating and destructive pathology in patients with multiple sclerosis (MS). We studied 100 patients with MS and 122 age, sex-, ethnic- and residence-matched controls. The DNA extraction and the genomic typing (A, B, DRB1 and DQB1 loci) were obtained with sequence-specific oligonucleotide method, using a commercially available reversible line blot assay (INNO-LIPA). All patients underwent a 1.5 tesla MRI examination of the brain. Disease severity was assessed by clinical (Expanded Disability Status Scale (EDSS)) and MRI (T2- and T1-lesion load (LL) and brain parenchymal fraction (BPF)) outcome measures. HLA-DQB1* 02 (OR 19.9, 95% C. I. 16.2-24.3, uncorrected (uncorr)- p<0.00001, corr-p<0.0006), -DQB103 (OR 16.8, 95% C. I. 13.6-20.5, uncorr-p<0.00001, corrp< 0.0006), -DRB115 (OR 4.6, 95% C. I. 3.7-5.6, uncorr-p= 0.0001, corr-p=0.006), and -DRB103 (OR 3.9, 95% C. I. 3.2-4.8, uncorr-p=0.0001, corr-p= 0.006) alleles were associated with MS. T2-, T1-LL, BPF and EDSS were not significantly different according to the carrier status of these HLA alleles. No differences were found in the ratios of disease severity/disease duration according to the HLA carrier status. Multiple regression analysis showed that a higher T2-LL was associated with the presence of DRB104 (uncorr-R2=0.15, p=0.006 and corr-R2=0.11, p=0.025) and B7 alleles (uncorr-R2=0.08, p=0.02 and corr-R2=0.07, p=0.018), T1-LL was associated with B7 (uncorr-R2=0.30, p<0.0001 and corr-R2=0.27, p=0.0001) and DRB112 (uncorr-R2=0.25, p<0.0001 and corr-R2=0.21, p=0.0002) alleles, whereas the BPF was predicted only by the presence of DRB112 allele (uncorr-R2=0.24, p=0.002 and corr-R2=0.20, p=0.004). The study findings suggest that some HLA alleles may predict the destructive pathological processes visible on MRI. Since the size of the sample studied is relatively small, further studies are needed to draw any firm conclusion about genotype/phenotype correlation in patients with MS.