Hierarchical Bayesian models (HBMs) have recently been advocated as formal, computational models of causal induction and reasoning. These models assume that abstract, theoretical causal knowledge and observable data constrain causal model representations of the world. HBMs allow us to model various forms of inferences, including the induction of causal model representations, causal categorisation and the induction of causal laws. It will be shown how HBMs can account for the induction of causal models from limited data by means of abstract causal knowledge. In addition, a Bayesian framework of the induction of causal laws, i.e. causal relations among types of events, will be presented. Respective empirical findings from psychological research with adults and children will be reviewed. Limitations of HBMs will be discussed and it will be shown how simple, heuristic models may describe the cognitive processes underlying causal induction. We will argue that formal computational model like HBMs and cognitive process models are needed to understand people's causal reasoning.