The present study investigated adult behavior while interacting with a three-month-old infant under conditions in which the child was introduced as a boy, as a girl, or with no gender information given. Gender labels did not elicit simple effects, but rather interacted significantly with the sex of the subject on both toy usage and physical contact measures. There was a stronger tendency for both male and female adults to utilize sex-stereotyped toys when the child was introduced as a girl. Most of the findings, however, reflected a differential response of men and women to the absence of gender information. In this condition, male subjects employed a neutral toy most frequently and handled the child least; in contrast, females used more stereotyped toys and handled the child more. All subjects attempted to guess the gender of the child (with boy guesses more frequent, although the child was actually female) and all justified their guess on the basis of stereotyped behavioral or physical cues like strength or softness.