The relaxation of land use restrictions is commonly cited as an effective solution to the housing affordability crisis because it would create more housing supply. However, one concern is that developers would not take advantage of such a policy reform because they behave dynamically, that is, they might delay construction in anticipation of higher returns in the future. How much of a problem is this practice of ‘‘land-banking’’? In this paper, we address this question by building a model of dynamic developer decision-making where developers choose when to build on a parcel of land. Using detailed microdata on the timing of housing projects in the Greater Toronto Area, we can estimate the extent to which developers engage in land-banking behaviour and what the key drivers of this behaviour are.