Relations between friendship (operationalized as reciprocated or nonreciprocated sociometric choices) and social competence were studied for children (mostly African American) attending Head Start. Initial analyses showed that children with reciprocated friends had higher social competence scores than children without reciprocated friends. Correlations suggested that the number of reciprocated friendships was associated with the social competence indicators studied here. Beyond the cost of having no reciprocated friends, having nonreciprocated friendships was not a liability. Cross-time analyses suggested differing patterns of relations for boys and girls. Having versus not having a reciprocated friend was unstable across time, because there was a trend toward participating in reciprocated friendships from 3 to 4 years of age (most older children had at least 1 reciprocated friend). For girls there was a positive relation between the number of reciprocated friendships at Time 1 and at Time 2. No benefit (in terms of social competence) was found for children making the transition from 1 classroom to the next with a friend.
