TECHNICAL difficulties have for a long time prevented any improvement in our knowledge of human leucocyte groups. The results of the Third Workshop on Histocompatibility Testing, held at Torino in 1967, showed that these difficulties are being overcome1. It now seems evident that the most important leucocyte antigens—transplantation antigens also present on platelets and most other tissues—belong to one very complex system (HL–A) controlled by the genes on at least two closely linked mutational sites on a pair of autosomal chromosomes2. This makes associations of antigenic factors possible, so that it is common to see the “inclusion” of some antigens in other “broader” antigens; that is, the antigen included is rarely or never present in individuals lacking the broad antigen3. This phenomenon has parallels in red cell serology; for example, the A1 antigen is included in the A antigen, and in the rhesus system, which has some similarities to the HL–A system, the C and D antigens are both included in the G antigen4. Within the HL–A system, however, such inclusions can often be explained by multispecificity of the serum used to define the broad antigen, and so absorption experiments are essential for the elucidation of this system. Absorption experiments have shown that the leucocytes from some individuals, although not agglutinated by a particular serum, nevertheless can completely absorb all agglutinating activity. Such cells are said to be agglutination-negative–absorption-positive (ANAP)5 and are considered to be positive for the antigen in question. A similar cytotoxic-negative–absorption-positive (CYNAP) phenomenon has been observed in some lymphocytotoxic antibodies, but some of these cases can be explained by the presence of anti-complementary factors in the serum6. We present here evidence that the “reaction-negative–absorption-positive” (RNAP) phenomenon and the inclusions can in some cases be explained by cross-reactive antibodies.