Eight disorders of the immune system have been described which are inherited as X-linked recessive conditions. The aim of this study has been to improve predictive testing in X-linked agammaglobulinaemia (XLA) and X-linked severe combined immunodeficiency (XSCID) and to investigate the underlying defect in XSCID. Precise genetic localisation is essential for accurate predictive testing and in order to develop strategies to clone the genes. Before disease localisation can be improved it is necessary to clarify the order of a number of probes within the region of interest. Fourteen genetic markers assigned to the X chromosome between the pericentromeric region and Xq22 were ordered by family studies and deletion mapping. Pulsed field gel electrophoresis was used to make a physical map of the markers linked to XLA. Knowledge of the order of these anonymous DNA probes led to finding additional linked probes for both diseases. This makes predictive testing possible in more families. There are X-linked and autosomal recessive forms of severe combined immunodeficiency. This has caused difficulties in counselling couples who have an affected male child and where there is no previous family history of the disease. In this study it has been shown that female carriers of the X-linked disorder have non-random use of the X chromosome in T lymphocytes. This provides a means of distinguishing between the autosomal and X-linked forms which enables more accurate genetic counselling. XSCID has been mapped to Xq11-q13 using DNA markers which detect polymorphic variation (de Saint Basile et al, 1987). No recombinations have been observed between the disease locus and the anonymous DNA probe cpX289. In this study the PGK1 locus was also shown to be closely linked to the disease. Using both of these linked markers predictive tests can be offered to 65% of families. The probe pSPT/PGK which detects a polymorphism at the PGK1 locus can also be used to investigate X chromosome usage. In females who are heterozygous for the polymorphism detected by this probe, carrier detection and assignment of phase can be carried out in the same procedure. This is a unique situation and is particularly useful when the proband could carry a new mutation or when there are no males available who can be used to assign linkage phase. It has been thought that XSCID results from a defect in a T lymphocyte specific gene because the phenotype is predominantly a lack of T lymphocytes and because host B lymphocytes produce functional antibody following transplantation and engraftment of T lymphocytes. Finding a non-random pattern of X chromosome usage in a mature cell population implies that the defective gene is expressed in that cell type and this technique was used to investigate gene expression. Non-random X chromosome usage was found in T lymphocytes, B lymphocytes, monocytes and granulocytes. The pattern of expression suggests that the underlying defect in XSCID is in a general metabolic pathway rather than a pathway specific to lymphocytes.