An experiment was addressed to three questions, (a) Is the retrograde amnesia (RA) phenomenon general enough to include a paradigm involving interoceptive stimuli? (b) Is RA due to a memory deficit, performance variables, or to the punishing effects of the RA treatments? (c) If RA is the result of a memory deficit, is it caused by erasure of the stimuli in memory or to disruption of the associative bond linking them? An amnesia agent (Metrazol) was administered in a discriminated taste-aversion paradigm, either within or after the CS-US interval. RA resulted in both cases, but was stronger when induced after the US. Due to the nature of the discriminated-avoidance paradigm used, the RA could not be explained in terms of the punishing effects of the RA treatment or general disinhibition. The data indicate that an amnesic treatment causes a memory deficit, at least part of which is due to the disruption of the OS trace.