The British Thoracic Association has recommended that close contacts of smear-positive cases of tuberculosis be followed up for at least 2 yrs (Tubercle 1978; 59: 245-259) but Selby et al. have recently suggested that a reduction in duration of follow up may be appropriate (Respir Med 1989; 83: 353-355). We have reviewed the results of contact procedures in Leeds to determine whether our experience supports reduction in the duration of follow up of contacts of patients with tuberculosis. In the 5-yr period 1983-87 there were 555 cases of tuberculosis (135 in Asians) of whom 42 (7.6%) were identified by contact procedures. In addition, contact procedures identified 35 children who were given chemoprophylaxis for positive Heaf tests (grade 2 or more). Of the 42 contacts with tuberculosis, 30 (71%) were diagnosed at the first visit, eight (19%) were diagnosed 6 months later and four (10%) were diagnosed 16-24 months after their initial clinic attendance. Five of the 42 contacts with TB were Asian, two of whom were diagnosed late. Seven out of ten non-Asian contacts who were diagnosed late had initial Heaf reactions of grade 1 or 2. All cases diagnosed late were contacts of a sputum-positive source. Poverty, as defined by residence in the Leeds Urban Priority Area, was associated with an increased risk of 3.3-fold for tuberculosis and a sixfold risk for chemoprophylaxis diagnosed by contact procedures.(ABSTRACT TRUNCATED AT 250 WORDS)