Eighteen of 67 patients who underwent cystometry for assessment of incontinence had a urinary tract infection 72 hours later. Though four of the 18 had an infection prior to cystometry, the true postcystometry infection was still high at 21%. Clinical details and urodynamic studies on these patients showed no correlation with sex, mobility, mental score, random blood sugar, renal function, initial residual volume of urine, previous pelvic operations or the type of bladder abnormality diagnosed on cystometry. The elderly and males with a high residual volume seemed more susceptible to infection. Thus cystometry carries a definite risk of infection even under optimal conditions and should not be undertaken lightly or without arrangements to follow up patients.