The authors investigate MacDonald's (1969) hypothesis that a sex by birth order interaction should be observed in measures related to sex typing. Since first borns are more thoroughly socialized, MacDonald reasons, and since parental expectations differ according to the child's sex, then first born boys and girls on sex typed variables. Forty-seven male and 46 female white, middle class, kindergarten children were measured on numerous sex typing indices: (1) a modified version of Brown's It Scale; (2) Draw-A-Person; (3) Toy Preferences; (4) Game Preferences; (5) Occupational Preferences; (6) Sex Role Adoption Ratings; (7) Sociometric Play Observations; (8) Child's Perception of Parental Dominance and Power; (9) Primary Mental Abilities; and (10) Peer Preferences. Hypothesized birth order differences were not observed for measures of sex role orientation, adoption, and preferences. Significant sex by birth order actions were observed on several less crucial measures. A modification of the hypothesis is thus suggested. (TL)