Methods used to evaluate subjective effects of electroconvulsive therapy (ECT) have relied on self-report about discrete aspects of memory. Although objective deficits are demonstrable, patients generally report improved memory after ECT. Patients have not been asked to evaluate the global impact of ECT on mood or memory. This study was undertaken to compare patients' evaluations of ECT outcomes using direct questioning about global impact compared with standard methods.A prospective, naturalistic study was conducted in seven hospitals. Patients completed clinical and neurocognitive assessments before ECT, including the Cognitive Failures Questionnaire (CFQ), the Autobiographical Memory Interview-Short Form, and a novel interview assessing expectations about the impact of ECT on mood and memory (Global Self-Evaluation-Mood [GSE-Md], GSE-memory [GSE-My]). Follow-ups were conducted one and 24 weeks after ECT, and the GSE-Md and GSE-My evaluated perceived global impact at these time points.Patients reported marked improvement after ECT on the CFQ, a traditional instrument assessing specific cognitive complaints. CFQ and depression severity scores were strongly correlated. On the GSE-My, patients reported a deleterious memory effect both one and 24 weeks after ECT. GSE-My, but not CFQ, scores were associated with treatment technique and long-term retrograde amnesia.Characterization of patients' experience of cognitive side effects after ECT differs markedly depending on assessment method. Direct questioning about global impact revealed more negative views and associations with objective indices of cognitive impairment. This represents the first report of concordance between subjective and objective measures of the effects of ECT on memory.