Electroconvulsive therapy (ECT) is an evidence-based treatment for difficult-to-treat, severe and sometimes life-threatening depressive episodes. Its effectiveness and cognitive side-effects are modifiable by stimulus dose, electrode placement and stimulus pulse width. 1 Regarding efficacy, meta-analyses of contemporary randomized controlled trials tell us that brief pulse high-dose unilateral ECT is like brief pulse bitemporal ECT, while ultrabrief pulse high-dose unilateral ECT is less efficacious. However, ultrabrief pulse high-dose unilateral ECT has fewer cognitive side-effects than brief pulse high-dose unilateral ECT, which has less than brief pulse bitemporal ECT. 2 In patients showing response to ECT, improvement is mostly witnessed early in the treatment course. In patients not responding after four to six treatment sessions, it is common practice to change the technique, either by switching electrode position, prolonging pulse width, or increasing the stimulus dose. However, remarkably, no high-level randomized controlled trial data are available to robustly support any of these clinical ECT practices or guide when to implement them. Instead, we have some observational data.In this issue, Hart et al. report the results of a large (N = 1699; 59.6% female, 91.8% white), decade-long, single-site, retrospective cohort study, identifying predictors of early and late response. 3 The cohort comprised patients who had a moderate-severe unipolar (78.8%) or bipolar depressive episode with a baseline Quick Inventory of Depressive Symptomatology (QIDS) score > 10. Applying various dosing protocols, patients were treated with thrice-weekly ECT, using mostly (94.1%) right unilateral electrode placement with an ultrabrief pulse (0.3-0.37 ms) electrical stimulus. One third (N = 555; 33%) of the patients met the response criterion (i.e., a decrease in QIDS score ≥ 50% from baseline) at treatment number 5, while nearly a quarter (N = 397; 23.4%) responded after treatment number 5. In early non-responders, changing the treatment technique from ultrabrief-pulse to brief pulse (i.e., >0.37 ms) unilateral ECT resulted in a higher likelihood of being a responder as compared to continuing ultrabrief-pulse ECT. This tantalizingly suggests that, for patients being treated with unilateral ultrabrief-pulse ECT, changing the treatment technique