SETTING: Thyolo district, rural Malawi. OBJECTIVES: To compare passive with active case finding among household contacts of smear-positive pulmonary tuberculosis (TB) patients for 1) TB case detection and 2) the proportion of child contacts aged under 6 years who are placed on isoniazid (INH) preventive therapy. DESIGN: Cross-sectional study. METHODS: Passive and active case finding was conducted among household contacts, and the uptake of INH preventive therapy in children was assessed. RESULTS: There were 189 index TB cases and 985 household contacts. Human immunodeficiency virus (HIV) prevalence among index cases was 69%. Prevalence of TB by passive case finding among 524 household contacts was 0.19% (191/100000), which was significantly lower than with active finding among 461 contacts (1.74%, 1735/100000, P = 0.01). Of 126 children in the passive cohort, 22 (17%) received INH, while in the active cohort 25 (22%) of 113 children received the drug. Transport costs associated with chest X-ray (CXR) screening were the major reason for low INH uptake. CONCLUSIONS: Where the majority of TB patients are HIV-positive, active case finding among household contacts yields nine times more TB cases and is an opportunity for reducing TB morbidity and mortality. The need for a CXR is an obstacle to the uptake of INH prophylaxis.