This paper contributes to the evaluation of property-led regeneration. Intervention in property markets remains an important strand in regeneration initiatives in Britain, and is still under-evaluated. In the mid 1990s, Robson and colleagues piloted a technique for helping to evaluate some of the effects of propertyled regeneration. This consisted of tracing so-called vacancy chains as the supply of new property stimulated turnover among users of buildings. Knowledge of the chains helps policy makers evaluate a particularly contentious aspect of area-based property-led initiatives: the extent to which they are stimulating new investment in a wider region, as opposed to simply moving existing employment around it. Robson's team undertook only a modest exercise in identifying chains (establishing around 40 chains in each UDC), and argued that further work was needed in order to establish the practicality of the approach and also to generate comparative data. This paper reports on a project that did this, tracing over 100 property chains in two parts of the Cardiff Bay Urban Development Area. The paper sets out, and discusses, the data acquisition techniques used, concluding that this is an approach that could have widespread application. It also reports on findings and concludes that the intervention by Cardiff Bay Development Corporation stimulated modest levels of new investment in Cardiff, but very little from beyond the sub-region.