The study of old fields has been integral to the development of succession theory. Some of the earliest studies of old fields were undertaken in the mid-1900s in northeastern North America, and these studies informed many of the classical ideas relating to old field succession. Today, long-term studies such as the Buell-Small Succession Study continue to provide important insights into vegetation development on abandoned farmland from both a succession theory and restoration practice perspective. It is of little surprise, therefore, that the theory of succession has had such success in describing community development on these sites, and those with similar climate, soils, vegetation, and legacy of land use. But how successful is succession for describing community development in other old fields around the world; those with harsher climates, poorer soils, different vegetation dynamics and more intensive land use? We use three case studies from temperate North America, tropical Latin America and the dry Mediterranean-climate area of southwest Australia to explore these questions. In particular, we consider how the plant traits of both native and introduced species interact with the legacy of land use to lead to successional trajectories that, at best, are unpredictable, and at worst, are stuck in a degraded state. How well does the theory of succession apply to ecosystems stuck in a particular state over time-scales beyond those of long-term research, and how does succession inform restoration in such systems? In conclusion, we present a generalised model of old field dynamics based on field studies from around the world, and discuss where succession is a success, where succession is mostly a success, and where other theories of vegetation development may better inform the restoration of old fields.