Telomerase activity assessment in voided urine is an important noninvasive tool for bladder cancer diagnosis. In a previous case-control study we verified that this method can detect tumor cells in urine with high sensitivity and specificity, but few data are still available on its accuracy to detect tumors in patients with symptomatic urinary tract diseases. Following recently published guidelines on bladder cancer, we aimed to define the diagnostic accuracy of urine telomerase levels in symptomatic patients. Telomerase activity, expressed in arbitrary enzymatic units (AEUs), was evaluated in urine collected from 515 patients: 197 with urinary tract symptoms and 318 with a first diagnosis of bladder cancer. Telomerase repeat amplification protocol (TRAP) sensitivity ranged from 93% to 61% and specificity varied from 42% to 88% at the different AEU cutoff values. At the cutoff of 50, the sensitivity was 87% (95% CI 83-91), the specificity was 70% (95% CI 63-75), and the overall accuracy, in terms of true positives and true negatives, was 80%. Sensitivity did not vary in relation to tumor grade or stage at diagnosis, or to patient age. Our results indicate that urine telomerase activity is a good marker for the early diagnosis of bladder tumors in symptomatic patients, a subset which represents an "at-risk" population requiring close surveillance.