Searching for an available parking space is a time-consuming and frustrating task for drivers, which also leads to huge amounts of fuel waste. Far from being solved, as the COVID-19 crisis slowly resolves, we can expect an increase in the use of private cars in comparison with public transportation, which will further aggravate the problem. Therefore, the use of smart parking solutions that help drivers to find an available parking space will become key to alleviate this problem. Moreover, existing solutions should be extended and adapted to incorporate the concept of social distancing, which is the only prevention measure (along with proper hygiene) which has been able to stop the virus so far. In this paper, we present an approach for the recommendation of parking spaces that considers social distance metrics to decide which parking spaces should be suggested to each driver. As far was we know, this problem has not been considered before in the literature. We illustrate the proposal with a use case scenario of a parking lot. Our experimental evaluation shows its feasibility and benefits.