This paper uses a difference-in-differences (DID) approach to identify the effect of proximity to COVID-19 cases on the returns of real estate firms. We use a novel micro level dataset which combines extensive data on the geographic footprint of COVID-19 patients, i.e. the locations they have resided in or visited, and the location of property holdings of real estate firms in Hong Kong. We find significantly negative effects of proximity to COVID-19 cases on stock returns. Having a property within 2 miles from a COVID-19 case results in a 0.02% lower return one day after the case disclosure. This effect is stronger for properties located closer and is weaker if the property is a residential building.