A large number of well-characterized HLA typing sera were used in a standard cytotoxicity technique to evaluate the frequency with which homozygous cells reacted to antisera directed against a crossreacting specificity. Four of 15 anti-HLA-A1 sera reacted with cells homozygous for HLA-A11 bet not with A11 heterozygotes. Similar dosage effects were noted with anti-A3 sera and A11 cells, with anti-A28 sera and A2 cells, with anti-A23 sera and A24 cells, and with anti-A24 sera and A23 cells. No dosage effects were seen with anti-A1 sera and A3 cells, with anti-A3 sera and A1 cells, and with anti-A11 sera and either A1 or A3 cells. Dosage effects were also not seen with anti-B51 or -B35 sera and cells containing antigens of the B5 crossreacting group. Dosage is a property of the individual serum and does not occur with all samples sharing the same primary specificity. As noted by red cell serologists, sera demonstrating dosage effects are common with some specificities but absent with others. The reactions noted may be a quantitative effect of epitopes present on certain antigens. Caution should be observed when interpreting an HLA phenotype that appears to contain two crossreacting antigens at the same locus, unless the sera used have been shown not to manifest dosage effects.