The Stan project is in development since 2011 and aims to enable efficient Bayesian inference. This tutorial will focus on the foundations of Stan, introduce the Stan modeling language, explain how to do Bayesian inference with Stan and finally address best practices. These will be introduced using examples of increasing complexity ranging from logistic regression to non-linear population pharmacokinetic/pharmacodynamic ODE models which will demonstrate the scalability and flexibility of Stan. Stan's key feature is the Hamiltonian MCMC sampler which is different than the various established flavors of Bayesian inference Using Gibbs Sampling (BUGS), such as WinBUGS, OpenBUGS, and JAGS. To fully exploit the advantages of Hamiltonian MCMC, participants will be briefly introduced to the foundations of Hamiltonian MCMC. After these more theoretical aspects, the Stan modeling language will be introduced. The Stan modeling language is inspired by the BUGS family such that BUGS users can quickly adopt Stan. Most importantly, participants will be taught best practices to write efficient Stan models. This will include how to debug Stan models easily and what to consider in order to expedite Stan models. These best practices will be presented using examples of increasing complexity. The examples presented will be run using the R package rstan.