OBJECTIVE: To investigate if women with recurrent urinary tract infection (UTI) warrant cystoscopy to exclude an abnormality of the lower urinary tract. This is particularly relevant given that non-invasive imaging has often been performed to exclude abnormality. Our further aims were to correlate imaging and risk factors with cystoscopic findings to determine their predictive value in finding pathology. PATIENTS AND METHODS: A database of women undergoing cystoscopy with recurrent UTI has been maintained at our institution for 10 years. We retrospectively examined this and patient records for patient demographics, and investigative and operative data. RESULTS: A total of 118 patients (mean 55 years) having recurrent UTI (mean 4.7 infections/year) were available. There were nine patients (8%) with significant abnormalities at cystoscopy: urethral stricture (six), bladder calculus (one), bladder diverticulum (one) and colovesical fistula (one). The negative predictive value (NPV) of imaging was 99% and significant (P < 0.01). Women with no risk factors for UTI had a NPV of 93% for normal cystoscopy (P > 0.05). The positive predictive value was low for imaging and risk factors in predicting cystoscopy findings. CONCLUSIONS: In our study, 8% of women had significant abnormalities detected during cystoscopy with most over 50 years. Women without risk factors for recurrent UTI and with normal imaging could have a cystoscopy omitted. Younger women are less likely to have pathology and this must be factored into decisions to perform cystoscopy.