<p>This paper describes a Petri net model, where information is attached to each token and when a transition fires, it can inspect and modify the information. The model is based on predicate/transitions (Genrich and Lautenbach) and on coloured Petri nets (Jensen).</p><p>This generalization of ordinary Petri nets allows, for many applications, more manageable descriptions, due to the fact that equal subnets can be folded into each other yielding a much smaller net. The paper investigates how to analyse high-level Petri nets, and it turns out that invariants and reachability trees, two of the most important methods for ordinary Petri nets, can be generalized to apply for high-level Petri Nets.</p>