Previous studies have shown that inducing elated and depressed moods by verbal mood inductions differentially affects the accessibility of memories of past happy and unhappy experiences. This study examined whether the accessibility effects obtained resulted from differences in moods induced or from differences in the procedures used to induce moods. Mood induction procedures from which all references to the subjects' life experience had been eliminated were used to induce elated and depressed moods. Subjects were then asked to retrieve past life experiences associated to each of a series of stimulus words. The experiences retrieved were subsequently rated for their original happiness/unhappiness and pleasantness/unpleasantness. Happy and pleasant experiences were more probable in elated than in depressed mood. Unhappy and unpleasant experiences were more probable in depressed than in elated mood. The extent of differences in accessibility between mood conditions correlated with the extent to which mood actually differed between conditions. There was no relationship between the extent of differences in accessibility and the extent to which subjects reported thinking about their life circumstances during the mood inductions. The results replicate previous findings and suggest that the observed differences in accessibility result from the differences in mood induced rather than directly from the differences in mood induction procedures.