OBJECTIVE: To assess the diagnostic accuracy of semen and urine culture in the diagnosis of chronic bacterial prostatitis (CBP). MATERIALS AND METHODS: In 70 consecutive men suspected of having chronic bacterial prostatitis along with 17 asymptomatic controls, we obtained urine and semen cultures followed 1 week later by the Meares and Stamey test, our reference standard. The interpretation of each of the cultures was blind to the results of other tests. RESULTS: 139 men were referred for evaluation of chronic bacterial prostatitis and 70 received all tests. Additionally, 17 control men volunteered to participate. The Meares and Stamey Test was positive in 69 (79%) patients. The semen culture had a sensitivity of 45% and a specificity of 94%. The likelihood ratio associated with a positive semen culture was 8.1 (95% confidence interval (CI) 1.2 to 55.3); the likelihood ratio associated with a negative semen culture was 0.6 (95% CI 0.5 to 0.7). The urine culture had a sensitivity of 4% and a specificity of 100%. The likelihood ratio of a positive urine culture was infinity and of a negative urine culture was 0.96 (95% CI 0.9 to 1). CONCLUSIONS: While a positive semen culture in a symptomatic patient may suffice to select and start antibiotic treatment against chronic bacterial prostatitis, a negative culture does not rule out the condition. Urine cultures alone are not useful for diagnosing CBP. The Meares and Stamey test remains important for the diagnosis of CBP in practice.