Primary iron overload may be relatively common in African Americans, but its cause is incompletely understood. Thus, we evaluated genotype and phenotype characteristics of unselected African American index patients with primary iron overload who reside in central Alabama. All had hepatic iron concentration > or =30 micromol/g dry wt or > or =2.0 g of iron mobilized by phlebotomy to achieve iron depletion. Genotype analyses were performed in African American control subjects from the same region. There were 23 patients (19 men, 4 women); mean age at diagnosis was 52 +/- 12 years (1 SD) (range 32-69 years). Nine (39.1%) reported that they consumed > or =45 g of ethanol daily; five had chronic hepatitis C. Eight had some form of hemoglobinopathy or thalassemia. Mean serum transferrin saturation was 56 +/- 28% (range 15-100%). The geometric mean serum ferritin at diagnosis was 1076 ng/mL [95% confidence interval 297-3473 ng/mL]. Increased stainable liver iron was observed in hepatocytes only in 4 patients, in macrophages only in 8 patients, and in hepatocytes and macrophages in 8 patients. The mean quantity of iron mobilized by phlebotomy (corrected for iron absorbed during treatment) was 5.3 +/- 2.0 g (range 4.0-8.4 g). Iron removed by phlebotomy was greater in patients with hemoglobinopathy or thalassemia than in those without these forms of anemia (6.6 +/- 1.3 g vs 3.9 +/- 1.6 g, respectively; P = 0.0144). Daily consumption of > or =45 g of ethanol or chronic hepatitis C was not associated with an increased or decreased amount of phlebotomy-mobilized iron, on the average. The percentage of index patients positive for HFE C282Y was greater than that of controls (P = 0.0058). The respective percentages of phenotype positivity for HFE H63D, D6S105(8), and HLA-A*03 were similar in patients and controls. HFE S65C, I105T, and G93R were not detected in index or control subjects. Two of 13 patients were heterozygous for the ferroportin allele nt 744 G-->T (Q248H), although the phenotype frequency of this allele was similar in patients and 39 controls. Synonymous ferroportin alleles were also detected in some patients. The ceruloplasmin mutation nt 1099C-->T (exon 6; Arg367Cys) was detected in 1 of 2 patients tested. Abnormal alleles of beta-2 microglobulin, Nramp2, TFR2, hepcidin, or IRP2 alleles were not detected in either of the 2 patients so tested. We conclude that primary iron overload in African Americans is not the result of the mutation of a single gene. HFE C282Y, ferroportin 744 G-->T, and common forms of heritable anemia appear to account for increased iron absorption or retention in some patients.