The purpose of this study was to ascertain whether and when temperament differences, including precursors of anxiety, emerge before onset and during stuttering development.The authors prospectively studied temperament characteristics of a community cohort of children who stutter (N = 183) and children in the control group (N = 1,261).No significant differences were found at ages 2, 3, or 4 years between children who stutter and control children for approach or at ages 3 or 4 years for easy/difficult temperament. Both of these measures are precursors of anxiety. Significant differences were found for reactivity and persistence at age 3 years. Children who stutter were less reactive to environmental stimuli and had a reduced ability to attend to a task until completion. There was no evidence of this difference for persistence at age 4 years. Reactivity was not measured at age 4 years.On the basis of parents' responses to the Short Temperament Scale, preschoolers who stutter did not have innately different temperaments from control children on those temperament traits measured from ages 2 to 4 years. They showed no signs of temperament precursors of anxiety before stuttering onset or shortly after. Results suggest, at most, that temperament is influenced somehow during the period after stuttering onset but with a waning developmental influence subsequently.