In their 1969 study, Osgood and Boucher proposed a hypothesis stating that positive words tend to be used more frequently and diversely than negative words in human communication. They called this the Pollyanna hypothesis. To prove their hypothesis, Osgood and Boucher studied 13 languages and found that across all the languages in question, the positive adjectives of Evaluative scales of the Semantic differential method were applied more frequently and used more diversely than their negative counterparts (with the exception of Finnish, which did not show more diversity on the usage of positive adjectives of the scales). It was also found that negative affixes (e.g., non-, un-, in English) were used more frequently to make positive adjectives of the scales (e.g., happy) into negative adjectives of the scales (e.g., unhappy) than the other way around. The present study tested if the Pollyanna hypothesis applied to the Japanese language by replicating Osgood’s and Boucher’s methodology. A word-association test was conducted with native speakers of the Japanese language as test subjects. Positive adjectives of Evaluative scales were used more diversely than their negative counterparts in the data obtained. The Pollyanna hypothesis was concluded to apply to the Japanese language diversity wise, but statistically significant results could not be obtained regarding frequency of usage.