This chapter starts with a discussion of Hume’s regularity theory of causality, which argues that the idea of necessary connection between two events (i.e., cause and effect) cannot be derived from observing the events and must be derived from an internal impression. Simply put, a causal relation is a mere constant conjunction of events. A counterfactual analysis of causation is based on Hume’s remark that if the cause did not occur, the effect would not exist. Lewis develops the analysis using the concept of possible world. In contrast to Hume’s deterministic view of causation, a probabilistic approach argues that causes raise the probability of − but do not necessarily lead to − their effects. Causal graph modeling and vector spacing modeling are two major techniques of identifying causal relations from empirical data, with the latter being more suitable for management research. A mechanism is a causal chain producing the effect of interest, and process tracing is a technique for identifying mechanisms in qualitative research.