The chapter gives an introduction to basic concepts of causal reasoning and to key methods for causal inference as developed and used in epidemiology over the last two decades. It begins with some considerations about formulating causal questions, e.g., using the idea of a target trial, and their mathematical formalization via potential outcomes. We then address typical assumptions and methods for estimating causal effects under a variety of causal models. The methods include estimating point- and time-dependent treatment effects, e.g., via inverse probability of treatment weighting, or using natural experiments such as instrumental variable estimation. While confounding bias is always a concern when using observational data, we discuss how the principle of target trial emulation offers some protection against other, often self-inflicted biases such as immortal time bias. The chapter concludes with the special topics of causal mediation analysis and causal discovery, which both aim at investigating direct and indirect causal pathways.