Quantitative precipitin studies with three anti-I sera show them to be distinctly different in specificity as revealed by their reactions with milk fractions, with Lea substances, with precursor and with cow blood group substances and with the various stages of periodate oxidation and Smith degradation of A, B and H substances. The various I specificities are concealed in interior structures of the blood group A, B, H, Lea and Leb substances and may be exposed by stepwise periodate oxidation and Smith degradation of A, B and H substances or by mild acid hydrolysis of B substance. Oligosaccharide inhibition assays showed one anti-I serum (Ma) to be inhibited best by a terminal nonreducing β-D-Gal-(1 → 4)-β-D-GNAc-(1 → 6)-structure present in a precursor blood group substance. From quantitative precipitin studies, the specificity of another anti-I serum was entirely different. Evidence suggesting that certain anti-I sera may not contain homogeneous populations of anti-I molecules is presented. The genetic implications of the findings are discussed. Anti-I eluted from OI erythrocytes behaved similarly to the anti-I in the original antiserum. I-anti-I specific precipitates dissolve readily at 37°C and reprecipitate rapidly on colling. The definition of the human I-anti-I system based on hemagglutination reactions in the cold may not have revealed its ture complexity especially with respect to its genetics.