A standardised procedure in contacts of patients with all types of newly diagnosed tuberculosis was undertaken by the British Thoracic Association Research Committee to assess the present relevance of contact examination in Britain. The results suggest that the tracing and examination of contacts remains a worthwhile procedure, resulting in the detection of significant numbers of previously unsuspected cases of tuberculosis in the contact population. The yield of new cases of tuberculosis is very similar for the Asian and non-Asian (mainly British) communities, namely 3.4% and 3.6% respectively, of the close contacts examined. The yield of new cases is about three times as great as the above percentages when the index case is positive on sputum smear, and about a third as great when the index case has non-respiratory tuberculosis. The examination of close contacts represents a larger workload in the Asian communities, where there are about five close contacts per index case, compared with about three in the other communities. Most close contacts were diagnosed at initial examination, but contacts of Asian index cases had an appreciable morbidity on re-examination at one or two years, as did the close contacts of smear-positive index cases of other ethnic groups. Prior BCG vaccination has a protective effect in both populations and chemoprophylaxis seems to be used infrequently in close contacts at high risk. All close contacts should be examined once. Close contacts of Asian index cases with respiratory disease, and close contacts of smear positive non-Asian index cases should be examined annually for at least two years; BCG vaccination or chemoprophylaxis should be considered in these groups. Casual contacts need be examined only if unusual exposure to a highly infectious case has occurred.