Another succinct, carefully qualified examination of contemporary research in the Developing Child series--this time focusing on siblings: whether sisters and brothers influence each other's development, why the tones of sibling relationships vary so widely, and what these observations about growing up together reveal about child development in general. Overall, there are fewer certainties here than from research on, say, early language or children's play. The complex patterns of mutual influence are difficult to observe and quantify, especially in the laboratory; studies have frequently highlighted the wrong variables; and relatively little research is available on siblings after the preschool years. Researchers have concluded that more easily controlled variables--birth order, age gap, sex--are less important (e.g., in explaining reactions to a new baby) than other factors: individual personality, relationship with parents before the birth, the mother's mood immediately after, and the sociability of the newborn. Also, psychologists seem to have repeatedly underestimated children's abilities. The often sophisticated capacities children demonstrate at home (for shared fantasies, recognitions of distress, opportunities for teasing) are significantly superior to the powers which experimental studies would suggest. While Dunn acknowledges that most siblings battle--with urgency, without inhibition--she believes that rivalry is neither the only nor necessarily the dominant feature in sibling relationships. She suggests that links between the development of emotion and the development of understanding are a more potentially productive area of inquiry, for these first sibling exchanges clearly color early social, emotional, and intellectual development. Parents won't find solutions to nightly fistfights or property disputes here, but, along with childcare professionals and child development students, they will find that this offers helpful ways of looking at problems even when many of the research results are tentative or incomplete.