BACKGROUND: Ultrabrief stimulation in electroconvulsive therapy (ECT) using a 0.25 or 0.30 ms pulse width markedly reduces the charge required to reach the seizure threshold (ST) and cognitive side effects. It is not known whether further reduction of pulse width to 0.15 ms is advantageous. METHODS: Thirty-seven patients were randomized to ST titration at the first session applying right unilateral (RUL) ECT with either a 0.15 or 0.30 ms pulse width and were titrated again in the second session using the alternative pulse width. All subsequent treatments used the pulse width applied in the second titration session, administering RUL ECT, starting at 6xST. The primary outcome was difference between the pulse widths in ST at the two titration sessions. Exploratory analyses examined differences in seizure duration and postictal time to recover orientation (TRO), averaged across all ECT sessions from the third onwards. Other exploratory analyses examined clinical improvement and retrograde amnesia for autobiographical information and other neuropsychological functions following the ECT course. RESULTS: In the first titration session, ST was significantly lower with the 0.15 ms than 0.30 ms pulse width. ST significantly increased when re-titrating with the 0.30 ms pulse width and significantly decreased when re-titrating with a 0.15 ms pulse width. There were no differences between the pulse width groups in clinical improvement, TRO, or neuropsychological measures. CONCLUSIONS: Ultra-ultrabrief stimulation with a 0.15 ms pulse width is more efficient in seizure induction than a 0.30 ms pulse width. Comprehensive studies should determine whether ultra-ultrabrief stimulation replaces ultrabrief stimulation as a default parameter for ECT.