Some models of cognitive development postulate qualitative change in the fundamental nature of mental structure, whereas others stress more invariant constraints on the form of mental organization. This article explores the implications of these two perspectives for characterizing the ontogeny of causal reasoning. Recent work on the development of causal reasoning over the preschool years has confirmed that children follow systematic constraints in identifying and labelling causes and effects. These observations have led many researchers to grant the preschooler an underlying knowledge of the defining principles of causality. This poses a challenge to the traditional view that the young child is precausal and must learn what features of occurrences distinguish causal from correlated events. The literature on causal reasoning in the preschool years is reviewed, and it is concluded that the hypothesis of an invariant causal scheme is only partially correct. Whereas preschoolers appear to share some principles of causal reasoning with adults, there are developmental changes in the extent to which these principles are held as necessary features of events, and in how they can be manipulated in the service of causal judgments and explanations. The issue of whether observed differences are due to change in specific knowledge or in the operational domain is raised, and it is argued that there is a reorganization in the interrelationship (although not the components) of the underlying causal scheme.