Clements (1962) has shown that atropine given intravenously has a vagal blockage effect superior to that achieved by subcutaneous injection. Electroconvulsive therapy is such a frequent reason for inducing general anaesthesia that a knowledge of the best premedication technique to use seemed to us important. Clearly the cardiovascular protective effect of the atropine is more important than the suppression of salivation. Maclay (1953) reported that in 62 deaths after E.C.T., 19 (30.6 per cent.) were attributable to the respiratory system. Clement's findings (1962) suggest that vagal blockade is adequate in protecting against cardiovascular disturbance 75 seconds after the injection of 1 mg. of atropine intravenously, but he mentions that the antisialogogic action is reputed to take longer to develop, and his paper did not include an assessment of this action.