Abstract
A small yet significant minority of contemporary patients with endogenous depressive illness who are treated with electroconvulsive therapy (ECT) gain little or no benefit. It is argued that the use of clinical features alone may not improve the ability to predict outcome after ECT. Many biological measures have been used to attempt to identify depressed patients for whom ECT would be an effective treatment, but none has yet been shown to be superior to clinical predictors. Depressed patients show a wide range of physiological responses to the first treatment of a course of ECT. Of these physiological responses, estimations of seizure threshold and of the release of posterior pituitary peptides merit further investigation as putative predictors of recovery.