Linear structural equation models have become indispensable in behavioral science research. They are a useful tool for testing hypotheses about relationships between variables and, even more importantly, for testing implications of causal structures. This book discusses the fundamentals, backgrounds, and procedures for specifying causal models. This is illustrated with an empirical example using the lavaan package within the R software. The focus of the book is less on the methodological and statistical aspects of a model and more on the adequate translation of theoretical concepts into a model. Special emphasis is placed on understanding the causal implications of models, the theoretical significance of variables, diagnosing a misfit to the data, and modern concepts.