BACKGROUND Surgical site infection (SSI) is a common post-operative complication causing significant morbidity and mortality. Many SSI occur after discharge from hospital. Post-discharge SSI surveillance in low and middle income countries needs to be improved. METHODS We conducted an observational cohort study in Dodoma, Tanzania to examine the sensitivity and specificity of telephone calls to detect SSI after discharge from hospital in comparison to a gold standard of clinician review. Women undergoing caesarean section were enrolled and followed up for 30 days. Women providing a telephone number were interviewed using a structured questionnaire at approximately days 5, 12 and 28 post-surgery. Women were then invited for out-patient review by a clinician blinded to the findings of telephone interview. RESULTS A total of 374 women were enrolled and an overall SSI rate of 12% (n = 45) was observed. Three hundred and sixteen (84%) women provided a telephone number, of which 202 had at least one telephone interview followed by a clinical review within 48 h, generating a total of 484 paired observations. From the clinical reviews, 25 SSI were diagnosed, of which telephone interview had correctly identified 18 infections; telephone calls did not incorrectly identify SSI in any patients. The overall sensitivity and specificity of telephone interviews as compared to clinician evaluation was 72 and 100%, respectively. CONCLUSIONS The use of telephone interview as a diagnostic tool for post-discharge surveillance of SSI had moderate sensitivity and high specificity in Tanzania. Telephone-based detection may be a useful method for SSI surveillance in low-income settings with high penetration of mobile telephones.