Birth weight is known to be a direct indicator of perinatal mortality and a clear predictor of adult pathologies too. It has been correlated with several causes of mortality in adulthood: low birth weight with diabetes, nephropathy and cardiovascular diseases and high birth weight with autoimmune diseases and cancer. In genome-wide studies, an extended human leucocyte antigen (HLA) region has been linked to birth weight variation. We focused our attention on the HLA haplotypes marked by HLA-A, HLA-B and HLA-DRB1 polymorphisms in 1206 healthy Caucasian newborns belonging to the Cord Blood Bank of Pavia (Italy) and their mothers, aiming to investigate the association between this restricted HLA region and birth weight variation. In our study, the HLA-B*38;DRB1*13 haplotype showed an ascending trend among centiles addressing to the high foetal weight. The HLA-A*02;B*15 haplotype showed a descending trend among centiles addressing to the low foetal weight. Besides the acknowledged correlation between the HLA-A*02 and HLA-B*15 alleles (as well as low birth weight) and type I diabetes and between the HLA-B*38 and HLA-DRB1*13 alleles (as well as high birth weight) and several autoimmune diseases, we cannot predict if our babies, healthy at birth, will suffer from these pathologies during life. Nevertheless, our data point to the HLA telomeric end for markers linked to the low birth weight and to the HLA centromeric end for markers linked to the high birth weight, thus limiting the region involved in birth weight variation, which still represents a useful predictor of disease risk in adulthood.