The relationship between the thyrotropin (TSH) response to thyrotropin-releasing hormone (TRH) and the duration of seizures induced by electroconvulsive therapy (ECT) in depressed patients was investigated. In a balanced-order cross-over design, 16 depressed women were given 0.4 mg TRH or placebo intravenously, 20 min before ECT in the first two sessions. In the third ECT session TRH was given just prior to ECT. Thyrotropin (TSH) levels at various sampling times, as well as the duration of seizures, were measured. There was a significant inverse correlation between plasma TSH concentrations 20 min after TRH administration (deltaTSH) and seizure duration. Furthermore, when patients were categorized according to their TSH response to TRH, the group with blunted responses (deltaTSH <6 microIU/ mL, n = 7) had a longer seizure time during ECT than the group with non-blunted responses (deltaTSH > 6 microIU/mL, n = 9). Finally, the seizure duration in the group with blunted TSH responses was reduced significantly when TRH was co-administered, while it remained unchanged in the group with non-blunted TSH responses. It is concluded that a blunted TSH response to TRH might indicate a seizure susceptibility as measured by the duration of seizures induced by ECT. The fact that TRH pre-administration had a reducing effect suggests that this substance might be involved in the pathophysiology of ECT-induced seizures.