provides improved analgesia in a number of settings. Their citation list is impressive, and I would note that systemic betamethasone has also shown promise. 2 Regarding their concerns about neurotoxicity, I agree that it is an unsettled issue. We made no assertions about the safety or recommendations for the use of perineural dexamethasone. One must not, however, be alarmist: a careful reading of one of their citations, an elegant preclinical model of adjuvant neurotoxicity, 3 reveals that, in the paradigm studied, ropivacaine was much more neurotoxic than dexamethasone.One larger question I would like to pose is, 'What is the duration of a nerve block?' Is it the duration of anaesthesia to pinprick? Is it the duration of motor weakness? Or, as I would argue, is it the length of time the patient is comfortable? In a real-world setting, we studied the most easily measured and clinically relevant outcome: how long patients were comfortable enough to not require other analgesics.Yes, systemic administration, if equally effective, would be much preferred. Including systemic administration in our trial would have required a much larger patient sample, as it was already a four-group trial. Consequently, we were forced to leave the question open. It will not, however, remain open for long.