In a factorial design, goldfish received 15 active-avoidance training trials on Days 1, 7, and 13 followed by electroconvulsive shock (ECS) or no treatment after the training session on Days 1 and 7. The final retention deficits observed when ECS was given after the f'trst session only or after the second session only were similar. Fish given ECS after both sessions showed a cumulative deficit that approximated the sum of the two effects. Similar results were obtained with puromycin. These data support the hypothesis that ECS and puromycin impair only recent learning. Retrograde amnesia gradients confirmed that the efficacy of these amnestic agents was not altered by a second administration. Intracranial injection of inhibitors of protein or of RNA synthesis to goldfish before or just following a training session disrupts fixation of long-term memory of the training session (Agranoff, 1974). Electroconvulsive shock (ECS) is also effective in producing amnesia (Davis et al., 1965), but the relationship of its mechanism of action to that of the antibiotics is unknown. Recent experiments with goldfish have indicated that ECS, like the protein and RNA synthesis inhibitors, can exert its effect proactively in a multitrial active-avoidance task (Springer et al., 1975). Since acquisition is demonstrated to be unimpaired, this result suggests that the amnestic effect of ECS is related to a sequela of the treatment rather than to events occurring at the time of shock administration, such as the resulting convulsions. A puzzling aspect of ECS effects in rats is the claim that memory can become refractory to disruption by repeated treatments (Nachman and Meinecke, 1969). If goldfish memory also becomes refractory to ECS the finding might well provide additional insights into the nature of ECS action. It was therefore of interest to investigate ECS impaired retention in goldfish following a second administration. The possible loss of effectiveness of a second puromycin injection after a second training session was also examined.