This paper addresses the following verification task: Given a graph transformation system and a class of initial graphs, can we guarantee (non-)reachability of a given other class of graphs that characterizes bad or erroneous states? Both initial and bad states are characterized by nested conditions (having first-order expressive power). Such systems typically have an infinite state space, causing the problem to be undecidable. We use abstract interpretation to obtain a finite approximation of that state space, and use counter-example guided abstraction refinement to automatically (iteratively) obtain suitable predicates for verification. Although our primary application is the analysis of graph transformation systems, we state our result in the general setting of reactive systems.