This reexamination of certain data in Study III of Johnson’s The Onset of Stuttering found that: (1) Among experimentals, children labeled as stutterers by at least one parent, at least 63% at onset time evidenced “primary stuttering” (simple repetitions and prolongations of sounds and syllables), and at least 28% evidenced only “normal nonfluencies” (repetitions of words and phrases, and other interruptions common in children), as indicated by parents' responses to one question. (2) This “primary stuttering” was much more frequent, and “normal nonfluency” much less frequent, in experimentals at onset of stuttering than in controls; and these differences were statistically significant. (3) Slight tension was the only “secondary” reaction indicated in more than 15% of the experimentals at onset, according to parents' responses to other questions. At onset experimentals and controls were significantly differentiated in incidence of tension, but not in incidence of indifference, awareness, or irritation. “Secondary” reactions in experimentals showed considerable increase within an average of about 18 months after apparent onset. This study supports the concept of primary stuttering as a beginning phase of a severity continuum.