To study social behavior in infancy, investigators usually focused their attention with the mother. But in recent years, considerable attention has been paid to the nature of infants' social interaction with their peers. A centrol issue underlying much of the early peer relational research has been the discovery of social interaction skills in infancy and the differences in mother-infant dyads and infant-infant dyads. The purpose of the study was to investigate infants' social behavior with peers versus mother. The issues raise in this study 1. Are there any differences or similarities between mother-infant and infant-infant dyads? 2. Are there any differences between two age group, and sex in social behavior their mother and peer? 3. Does the infant’s social behavior affected by sex of partner, infant’s birth order and time of play? The subjects were 48 normal home-reared children drawn on the basis of age between 10-13 months and 22-25 months. The 20 minutes of semi-structured experimental sessions were videotaped and coded on the checklist which was developedby the researcher. The data was analyzed by frequency and percentage, one-way analysis of variance, two-way analysis of variance, and trend analysis with the SAS computer program. The findings of this study are as follows : 1. Infants interacted differently toward their mother and peer. Watching was occurred most frequently in infant-infant dyads, and vocalization was in mother-infant dyads. Behaviors such as facial expressions, vocalization, physical contact are most frequently occurred between mother-infant dyads. Interactions with same play materials and exchanging toys were more frequently observed in infant-peer dyads. Both the frequency of social behaviors with the peer and the mother increased reliably with age. 2. Two--way analysis of variance on the social behavior with age and sex were analysed. Watching the mother indicated a reliable change in behavior with sex. and watching the stranger indicated a reliable change with age. Vocalizing to mother & peer were increassed with age, but monologing during play was decreased. Facial expressions indicated a reliable change with age. Smiling & laughing were more frequent than crying and fussing. Infants were more likely to approach to a peer, but at the same time more likely to touch their mothers. Approaching to peer was increased and touching their mother decreased with age. Solitary play occurred in similar frequency in 10-13 month and 22-25 month groups. Social play, in contrast increased reliably with age. 3. Infants were participated more in social play to same-sex dyads than to opposite-sex dyads. But there were no reliable effects of sex in watching, vocalizing, facial expression, and physical contact. Birth-order differences in social behavior had no reliable effect. The effect of time on social behavior yieled no differences but watching behavior was decreased with time. And the trend of watching behavior was linear-shaped.