To test the commonly held premise that prematurity is a risk factor for problems of behavior and social interaction, as well as cognitive and physical development, temperament and behavior of children born preterm and full-term control subjects were compared at five separate time periods from infancy through early school age. All the preterm infants (n = 126) of a representative group of infants enrolled in a longitudinal study (n = 2443) were surveyed at 4 to 8 months corrected age. Subsamples of the group were studied further in successive years as young toddlers (n = 65), older toddlers (n = 60), preschoolers (n = 84), and at early school age (n = 81). At each period there were no differences between those studied and those not studied on socioeconomic status, gestational age, sex, or birth order. Parental ratings of temperament and behavior were used at appropriate ages: the Revised Infant Temperament Questionnaire, the Toddler Temperament Scale, the Childhood Temperament Questionnaire, Behaviour Checklist, Preschool Behaviour Questionnaire, and Rutter's Childhood Behaviour Questionnaire (CBQ), as well as mother's overall rating of temperament. In infancy there were no significant differences on temperament dimensions, clinical temperament categories, or parental ratings of individual or composite behaviors between the preterm and full-term groups. For toddlers, temperament scores were similar for the two groups but preterm subjects were significantly more likely (P less than .01) to have an easy temperament and less likely to have a difficult temperament. There were no differences on any of the other temperament or behavior ratings.(ABSTRACT TRUNCATED AT 250 WORDS)