An Autobiographical Memory Interview (AMI) was administered to 75 depr
essed inpatients and 16 nondepressed controls. Patients were randomize
d to 1 of 4 forms of electroconvulsive therapy (ECT) that varied in el
ectrode placement and stimulus intensity. Short-term retrograde amnesi
a was assessed during the week following the randomized phase. Bilater
al ECT produced more marked deficits than right unilateral ECT. At a 2
-month follow-up, persistent amnesic deficits were related to having r
eceived a second ECT course and, to a lesser extent, bilateral ECT dur
ing the randomized phase. The magnitude of clinical improvement was no
t associated with amnesia scores at either time point. There were no d
ifferential amnesic effects as a function of the affective valence of
memories. It appears that retrograde amnesia for autobiographical info
rmation after ECT and mood congruence effects on recall are independen
t phenomena. The magnitude and persistence of retrograde amnesia is re
lated to how ECT is performed and not to changes in clinical state or
the affective valence of memories.