Background. This study investigated the patient's self-report and clinician's confirmation of tracheoesophageal voice prosthesis leakage patterns (through or around) with or without the cough reflex and whether prosthesis diameter affected the leakage route. Methods. Sixty-six consecutive participants with a total of 200 patient-initiated reasons for prosthesis changes were en- rolled prospectively. Patient's self-report of leakage and cough reflex were recorded prior to clinician's confirmation. Results. One-hundred eight (54%) of the 200 patient-initi- ated reasons for prosthesis changes were leakage through or around the voice prosthesis. Leakage was unrecognized in 21 (23%) of 92 instances, even though 15 (71%) of those 21 instan- ces exhibited a cough reflex. Clinician's confirmed leakage in 118 (59%) of 200 patient-initiated reasons for prosthesis changes. Coughing occurred significantly less with leakage around (9 (53%) of 17 instances) than that with leakage through