Routine laboratory screening is based on the detection of WNV specific IgM and IgG in blood and cerebrospinal fluid. Confirmation is then classically applied by real timeRT-PCR (rRT-PCR)in Cerebrospinal fluid (CSF), which often gives negative results due to too short virorachia and late sampling.rRT-PCR was applied-for the first time for routine diagnosis purpose-on urine samples. During 2018 outbreak in Tunisia, 107 patients presented WNV neurologic symptoms and were positive for WNV serology. Of them, 95 patients were sampled for urine and 35 were sampled for CSF. Qualitative rRT-PCR was performed on both type of samples. WNV RNA was detected in 50.5% of urine samples (48/95) and in 2.8% of CSF samples (1/35). WNV RNA was detectable from day1 to day41from symptom onset, however, positive urine rate was 53.1% during the first 10days from symptom onset. The proportions of urine-positive and urine-negative samples, based on day of collection, showed no statistical difference (p > 0.005). Cycle threshold (Ct)values ranged from 12 to 39, with no correlation with the day of collection. The lowest Ctvalue was detected for urine sampled on day 5 after symptom onset. A statistically significant difference was found between age groups of confirmed and non confirmed cases (p < 0.001). Our study reported the use of rRT-PCR on urine samples as a confirmatory diagnostic tool for WNV "probable cases" during an outbreak. Our findings underlined the reliability and therapidity of this confirmatory tool, even late, and showed its superiority on CSF investigation.