The prevalence of tuberculosis in the household contacts of 67 smear-positive patients was compared with that in a 'non-contact' population. The prevalence of smear-positive cases among adult contacts was 3.03% as opposed to 1.35% in non-contacts; 27.9% and 11.43% respectively of children in the two groups had positive tuberculin tests (Heaf grade III or IV). The differences in children are statistically significant and could be used to justify the contention that active case-finding in underdeveloped areas should be limited to intensive contact-tracing. It is argued, however, that this may be an inherently conservative response that will not meet the needs of tuberculosis control. A plea is therefore made for a redistribution of resources in order to meet these needs.