In the annals of psychiatric treatment, the advent of Cardiazol therapy has been afforded merely passing mention as a stepping-stone to the development of electroconvulsive therapy. Yet in the 1930s it was the most widely used of the major somatic treatment innovations in Britain's public mental hospitals, where its relative simplicity and safety gave it preference over the elaborate and hazardous insulin coma procedure. Devised on a dubious hypothesis of biological antagonism, Cardiazol armed psychiatry with an immediately effective weapon in the battle against schizophrenia, an enduring and debilitating condition responsible for over half of the mental hospital population. What made Cardiazol work – or appear to work? This account shows how evaluation of convulsive therapy was skewed by naÔve outcome measurement and diagnostic discrepancies, and how its therapeutic indication evolved from schizophrenia to affective disorders. Psychological mechanisms are considered, with the suggestion that the intense fear experienced during treatment – the major reason for abandoning Cardiazol in favour of electroshock – was therapeutically advantageous.