Electroconvulsive therapy is an effective treatment for severe and medication-resistant depression. There have been no reports describing how a volatile anaesthetic affects haemodynamic responses, seizure duration, and recovery characteristics during electroconvulsive therapy. We carried out a repeated-measure crossover study to compare the effects on haemodynamic responses, seizure duration, and recovery characteristics of the following types of anaesthesia in electroconvulsive therapy: propofol alone, sevoflurane alone, and propofol combined with sevoflurane. We recruited 50 patients requiring electroconvulsive therapy for depression. For anaesthesia induction, 1.5 mg/kg propofol (condition P), 5% sevoflurane in oxygen following a vital capacity rapid inhalation induction (condition S), or 1.5 mg/kg propofol followed by 5% sevoflurane in oxygen (condition PS) was administered. Succinylcholine 1.5 mg/kg was then given. Electrical stimulation was administered after fasciculation. Measurements were obtained before anaesthesia induction (baseline), prior to succinylcholine administration, prior to electroconvulsive therapy, and at the peak after electroconvulsive therapy. After electroconvulsive therapy, peak heart rate and peak mean arterial pressure were highest in condition S. Whereas recovery time was longest in condition PS, motor seizure duration was significantly shorter than in either condition P or S. Electroencephalographic seizure duration was significantly shorter in condition PS than in condition P and significantly shorter in condition S than in condition P. Sevoflurane anaesthesia alone is most disadvantageous in terms of haemodynamics. Propofol-sevoflurane anaesthesia is advantageous in terms of haemodynamics, but disadvantageous in terms of seizure duration and recovery time. Propofol alone is most advantageous in terms of seizure duration.