Since electroconvulsive therapy (ECT) can result in generalized seizures that lack efficacy, physiological markers of treatment adequacy are needed. Specific electroencephalographic (EEG) features differentiate seizures produced with barely suprathreshold right unilateral (RUL) ECT, an ineffective treatment, from effective forms of ECT. This study determined whether EEG features are sensitive to treatment condition using a broad dosing range for RUL ECT, as well as predictive of clinical and cognitive outcomes. Quantitative EEG measures and observer ratings were compared in predictive power. From a larger study, 54 in-patients with major depression were randomized to low (1.5 x seizure threshold (ST)), moderate (2.5 x ST), or high-dose (6 x ST) RUL ECT, or high-dose (2.5 x ST) bilateral (BL) ECT. High dosage RUL and BL ECT were comparable in efficacy, and superior to low and moderate dosage RUL ECT. In the slow frequency bands (delta), BL ECT resulted in greater ictal power, ictal coherence, and postictal suppression than each RUL ECT condition, but the EEG measures failed to discriminate the RUL ECT groups. EEG measures were modestly associated with clinical outcome, with greater ictal power, delta coherence, and postictal suppression positive predictors. None of the EEG measures were associated with cognitive outcomes. Inability to distinguish forms of RUL ECT differing markedly in dosage and efficacy suggests that EEG measures have limited potential as markers of treatment adequacy. Rather than assaying treatment adequacy, the EEG features associated with efficacy may reflect individual differences in the strength of inhibitory processes that terminate the seizure, and can help isolate the biological variability that predisposes to positive or negative clinical response to ECT.