In the first nationwide community-based survey of the epidemiology of tuberculosis in Saudi Arabia, 7721 subjects were screened in the 5 provinces (using an equal proportional allocation formula) for 2 parameters: (1) prevalence of positive Mantoux test in non BCG vaccinated subjects; (2) prevalence of bacillary cases on sputum culture. The prevalence of positive Mantoux reaction in children aged 5-14 years was 6% +/- 1.8; higher in urban areas (10%), and lower in rural areas (2%), thus classifying Saudi Arabia among the middle prevalence countries. These relatively good results (by Third World standards) could reflect the rise of the standard of living and wide availability of free treatment for active cases with a lowered risk of infection in the community. This view is supported by the fact that in our survey, only one subject grew Mycobacterium tuberculosis in the sputum. However, there were foci of high prevalence of Mantoux reaction in the urban communities in the Western province (20% +/- 8.7 urban; 1% +/- 1.9 rural). The problem may be caused by the fact that the province receives every year over a million pilgrims, some of whom are known to settle illegally and escape the usual screening for tuberculosis imposed on foreign labourers. In conclusion, even in the absence of an enforceable national programme for the eradication of tuberculosis, the economic standard and wide availability of free treatment for active cases has resulted in relatively low rates of prevalence of tuberculin sensitivity in children. The foci of high prevalence in the Western Province require special screening arrangements.