A population-based study in South East Wales (population 668,100) identified 69 families with 135 affected members with von Recklinghausen neurofibromatosis (NF-1), giving a disease prevalence of 20/105 of population. In these families penetrance of the NF-1 gene was 100% by the age of five years. 41/135 cases were judged to represent new disease mutations and the mutation rate was estimated to lie between 3.1x10-5 and 10.4x10-5. A parental age effect for new mutations was not demonstrated, nor was a maternal effect on disease severity. The clinical features and natural history of NF-1 in this cohort were used to derive data for genetic counselling and recommendations for the management of affected individuals. For counselling purposes the complications of NF-1 can be usefully divided into 4 categories (the frequency of each, based on this study, are shown in parentheses): intellectual handicap (33% overall, moderate/severe retardation 3.2%, minimal retardation/ learning difficulties 29.8%); complications developing in childhood and causing lifelong morbidity, e. g. facial plexiform neurofibromas, scoliosis, pseudoarthrosis (8.5%); 'treatable' complications which can develop at any age, e. g. benign disorders of the nervous system, visceral and endocrine tumours, renal artery stenosis (15.7%) and malignant or CNS tumours (4.4-5.2%). The study population indicates that sufferers are not being diagnosed sufficiently early, nor receiving appropriate follow-up and counselling. It is recommended that patients with NF-1 have regular clinical assessments to monitor for the development of complications, although none occur often enough to warrant biochemical or radiological screening. As many of the complications develop early in life, children should have biannual review; in adults, unless a particular complication indicates more frequent review,annual clinical examination is sufficient. Alongside the population survey, genetic linkage studies were undertaken in selected large families to determine the chromosomal localisation of the NF-1 gene. At the outset of this work, two families had been reported in which NF-1 and Myotonic Dystrophy (DM) appeared to co-segregate, suggesting that the two genes were closely linked and on chromosome 19. However, linkage studies of 3 chromosome 19 markers linked to DM showed significantly negative lod scores, therefore excluding this possibility. Other chromosomes were then studied using random unique sequence DNA probes and samples from the largest families were made available to collaborators in the USA for linkage studies using possible candidate genes (ß nerve growth factor and oncogenes). No marker studied showed evidence of linkage. The negative data were used to produce an exclusion map for NF-1, using the computer program 'EXCLUDE'. The presentation of this work was one of the factors which precipitated the formation of an international consortium for NF-1 linkage in February 1987; the first task of the consortium was to produce an expanded exclusion map. A small positive lod score for a marker on chromosome 17, taken with the exclusion data, showed that NF-1 was seven times more likely to be on chromosome 17 than any other chromosome; this was rapidly confirmed by two North American groups, one of which was using samples from the 5 largest families presented in the thesis. Subsequent linkage analysis of pericentromeric chromosome 17 markers in the Welsh family panel showed no evidence of non-allelic heterogeneity and identified closely linked flanking markers for the NF-1 gene suitable for prenatal/presymptomatic diagnosis. The chromosomal localisation of NF-1 represents a major step towards the eventual understanding of the disease pathogenesis and the development of possible treatments.