The authors conducted 9 experiments to test the hypothesis (S. Schwartz, 1975) that arousal influences the accessibility of information stored in memory. They investigated the relationship between arousal levels (as indexed by personality types) and the type of stimuli or cues presented during study or test. They predicted that low-arousal individuals (stable extraverted individuals in Experiments 1-3 and 5-9 and high-impulsive individuals in Experiment 4) would be influenced by semantic stimuli, whereas high-arousal individuals (neurotic introverted individuals in Experiments 1-3 and 5-9 and low-impulsive individuals in Experiment 4) would be influenced by physical (i.e., graphic, phonetic, or both) stimuli. They tested the arousal-accessibility hypothesis by using a variety of tasks including verbal discrimination, false recognition, cued recall, and paired associates. With the exception of the finding that stable extraverted participants performed better than neurotic introverted participants on an incidental associative-matching task (Experiment 3), the results from the verbal discrimination studies (1-5) did not support the hypothesis. In Experiment 6, the authors tested the hypothesis by using a false-recognition task. False alarms varied as a function of phonetic and semantic stimuli, but personality types were not differentially sensitive to the manipulation. The same was true for the cued-recall studies (Experiments 7 and 8); personality types were not differentially sensitive to the semantic and phonetic stimuli. Experiment 9 (paired-associate learning) was a replication of Schwartz's study. The authors found some support for the Schwartz hypothesis: Extraverted participants were adversely affected by semantic similarity. Overall, the findings did not provide much support for the arousal-accessibility hypothesis.