We provide a formalisation of the theory of pushdown automata (PDAs) using the HOL4 theorem prover. It illustrates how provers such as HOL can be used for mechanising complicated proofs, but also how intensive such a process can turn out to be. The proofs blow up in size in way difficult to predict from examining original textbook presentations. Even a meticulous text proof has “intuitive” leaps that need to be identified and formalised.