Access to urban databases is a major issue regarding urban climate research. The most useful databases are the ones that demonstrate both high quality and large spatial coverage. OpenStreetMap is a major geographical data project that spans worldwide. In this crowdsourcing project, data is partly gathered by users. Thanks to its open-source architecture, OpenStreetMap shows promising potential as a resource for urban climate informatics. However, it appears that information relevant to urban climate research is sometimes missing in cities, notably building height. The issue of missing building heights has been addressed within GeoClimate’s framework initiative. Geoclimate is an open-source geospatial toolbox that computes Urban Canopy Parameters (UCPs) at various scales as well as Local Climate Zones. As part of the GeoClimate project, a methodology has been proposed to estimate the missing building heights in OpenStreetMap. This approach uses a random forest algorithm to calculate building heights based on 62 UCP. It has been previously tested on 22 French cities. However, it needs to be evaluated under different urban morphologies and land use configurations. This work applies the building height estimation algorithm for Hong Kong. This city hosts a high density of high-rise buildings that strongly differs from French cities’ urban form. Building heights calculated using GeoClimate are compared to building heights provided by reference databases and the main differences are further investigated.