Caffeine lengthens electroconvulsive therapy (ECT) seizures but has not been demonstrated to lower seizure threshold, nor to directly improve treatment outcome. We examined the effects of caffeine upon three recently proposed physiologic measures of seizure quality or impact: electroencephalography voltage suppression ratios, seizure regularity, and heart rate. None of these measures differed significantly in depressed inpatients randomly pretreated with either i.v. caffeine or saline on successive treatments during a course of right unilateral ECT. Newly defined measures of seizure impact may help identify ECT seizures of differing clinical efficacy, but these measures did not discriminate our caffeine and placebo conditions.