Before the discovery of HFE, reports suggested that hemochromatosis patients with the ancestral haplotype (or some element thereof) have more severe iron overload than those without the haplotype. We performed univariate and multivariate analyses of the relationships of human leukocyte antigen (HLA)-A*03 and HLA haplotype A*03-B*07 to iron measures (serum iron concentration, transferrin saturation, and serum ferritin concentration at diagnosis and units of phlebotomy to achieve iron depletion) in hemochromatosis probands homozygous for HFE C282Y diagnosed in medical care. Iron overload was defined by demonstration of hepatic iron index of ≥1.9 or removal of ≥2.0 g Fe by therapeutic phlebotomy. We tabulated the phenotype frequencies of HLA-A*03 and the frequencies of common HLA haplotypes A*01- B*08, A*02-B*44, A*03-B*07, and A*03-B*14 in three groups of white adults: (1) 141 hemochromatosis probands with C282Y homozygosity; (2) 195 index cases with IgG subclass deficiency (IgGSD) or common variable immunodeficiency (CVID), disorders typically linked to Ch6p, and (3) 750 control subjects. Among probands, 86 men and 42 women had iron overload. Frequencies of HLA-A and -B alleles in probands did not depart significantly from Hardy-Weinberg equilibrium. The phenotype frequency of A*03 did not differ significantly between men and women in the each of the respective three groups. The frequency of haplotype A*03-B*07 was greater in men than women with hemochromatosis (0.3081 vs. 0.1455; P = 0.0019). The frequency of A*03-B*014 was significantly greater in women than men with hemochromatosis (0.1182 vs. 0.0407, respectively; P = 0.0134). Mean values of most iron measures were not affected by numbers of copies of A*03 or by presence of A*03-B*07 in either men or women in univariate analysis. ANOVA models of sex, age at diagnosis, and all HLA alleles and haplotypes in probands were used to determine effects of these variables on iron measures. ANOVA models revealed that (1) there were no significant predictors for serum iron concentration; (2) B*14 is associated with higher transferrin saturation in women and lower transferrin saturation in men; (3) A*01-B*08 is associated with a trend of higher serum ferritin levels; and (4) A*03-B*14 is associated with exaggeration of the age-associated upward trend in units of phlebotomy to achieve iron depletion. In hemochromatosis probands with HFE C282Y homozygosity, we conclude that (1) disparate frequencies of HLA haplotypes A*03-B*07 and A*03-B*14 occur in men and women and (2) HLA-A*03 and HLA-A*03-B*07 are not independent variables associated with iron overload severity. © 2004 Elsevier Inc. All rights reserved.