The syndromic diagnosis of sexually transmitted infections (STIs) is widely recognized as the most practical, feasible, and cost-effective diagnostic tool in resource-limited settings. This study assessed the diagnostic accuracy of syndromic versus laboratory testing of STIs among 794 men randomly selected from the Moshi district of Tanzania. Participants were interviewed with a questionnaire that included questions on history of STIs symptoms. Blood and urine samples were taken from the participants for laboratory testing. Only 7.9% of the men reported any symptoms of STI; however, 46% of them tested positive for at least one STI. There was little agreement between syndromic and laboratory-confirmed diagnoses, with low sensitivity (0.4%–7.4%) and high specificity (96%–100%) observed for each individual symptom. The area under the receiver-operating curve was 0.528 (95% CI: 0.505–0.550), indicating that the syndromic approach has a 52.8% probability of correctly identifying STIs in study participants. In conclusion, whenever possible, laboratory diagnosis of STI should be favored over syndromic diagnosis.