Objectives: The household is important in influenza transmission due to intensity of contact. Previous studies reported secondary attack rates (SAR) of 4-10 for laboratory-confirmed influenza in the household. Few have been conducted in middle-income countries. Methods: We performed a case-ascertained household transmission study during May-October 2013. Index cases were patients with influenza-like-illness (cough and self-reported or measured fever (=38 ?C)) with onset in the last 3 days and no sick household contacts, at clinics in South Africa. Household contacts of index cases with laboratory-confirmed influenza were followed for 12 days. Results: Thirty index cases in 30 households and 107/110 (97 ) eligible household contacts were enrolled. Assuming those not enrolled were influenza negative, 21/110 household contacts had laboratory-confirmed influenza (SAR 19 ); the mean serial interval was 2.1 days (SD = 0.35, range 2-3 days). Most (62/82; 76 ) household contacts who completed the risk factor questionnaire never avoided contact and 43/82 (52 ) continued to share a bed with the index case after illness onset. Conclusion: SAR for laboratory-confirmed influenza in South Africa was higher than previously reported SARs. Household contacts did not report changing behaviors to prevent transmission. These results can be used to understand and predict influenza transmission in similar middle-income settings.