Graft-versus-host disease (GvHD) after bone-marrow transplantation in dogs is controlled by many different genetic systems. In littermate combinations identical for the major histocompatibility complex (MHC) the number of systems that influence GvHD is related to the number of donor lymphocytes injected. If the number of donor lymphocytes administered is sufficiently low, minor histocompatibility systems do not influence survival after bone-marrow transplantation. With increasing numbers of donor lymphocytes the beneficial influence of MHC matching on GvH incidence and severity disappears and minor histocompatibility antigens, coded for on at least two other autosomal chromosomes as well as possibly the Y chromosome, can cause severe GvHD. In contrast, the X chromosome does not appear to carry a histocompatibility system that is of relevance to GvHD control. The severity and tissue distribution of histological signs of GvHD in recipients of bone-marrow and lymph-node cells from MHC-identical donors are similar to those in recipients of MHC-mismatched bone-marrow cells. Female donors do appear to cause severe GvHD more frequently than males. In contrast to rhesus monkey and human bone-marrow cells, dog bone-marrow cells are negative in PHA tests. This is in accordance with the generally benign course of GvHD in dogs that are treated with bone-marrow cells only from histocompatible littermate donors. The influence of the sex of the bone-marrow donor on GvHD incidence and severity is not reflected in differences between PHA tests with male and female dog lymphocytes. A better predictive test for GvH potential than the PHA test appears to be needed. Alternatives to additional donor selection for the prevention of GvHD in histocompatible recipients appear to be the use of a male donor and the removal of lymphocytes from bone-marrow-cell suspensions prior to transplantation.