To assess the clinical value of human leukocyte antigen typing in the diagnosis and management of hereditary hemochromatosis, 105 siblings of 35 proband cases of hemochromatosis were retrospectively analyzed to study whether the exclusion of human leukocyte antigen typing would have adversely affected management. All siblings and probands had already been tested for human leukocyte antigen-A and human leukocyte antigen-B typing, serum ferritin and transferrin saturation. The median age of siblings was 55 yr (range = 11 to 82). Siblings were categorized according to putative genotype (homozygote, heterozygote and normal) using human leukocyte antigen typing. Phenotypic expression of hemochromatosis was considered to be iron overload as indicated by an elevated ferritin (male = greater than 350 micrograms/L, female = greater than 200 micrograms/L) and/or transferrin saturation (greater than 55%). Six of 37 homozygotes had a normal ferritin and transferrin saturation, with five of these patients under 32 yr old. No putative heterozygotes with both an abnormal ferritin and transferrin saturation were seen, although 12 of 48 (25%) heterozygotes had either an elevated ferritin or transferrin saturation. Twenty of 20 normal siblings had a normal ferritin and transferrin saturation. To assess the cost of screening with and without human leukocyte antigen typing, a cost model simulation was used that compared the costs of both methods in a hypothetical family (proband, homozygote, heterozygote and normal sibling).(ABSTRACT TRUNCATED AT 250 WORDS)