For researchers using observational data, a useful way to answer a causal question is to design the target trial that would answer it and then emulate its protocol. The example of the HIV-treatment-as-prevention strategy illustrates the benefits of this approach.