In a scenario where there is no vaccine for COVID-19, non-pharmaceutical interventions are necessary to contain the spread of the virus and the collapse of the health system in the affected regions. One of these measures is social distancing, which aims to reduce interactions in the community by closing public and private establishments that involve crowds of people. The lockdown presupposes a drastic reduction in community interactions, representing a more extreme measure of social distancing. Based on geolocation data provided by Google for six categories of physical spaces, this article identifies the variations in the circulation of people in South America for different types of social distancing measures adopted during the COVID-19 pandemic. In this study, population mobility trends for a group of countries between February 15, 2020 and May 16, 2020 were analyzed. To summarize these trends in a single metric, a general circulation index was created, and to identify regional mobility patterns, descriptive analyzes of spatial autocorrelation (global and local Moran index) were used. The first hypothesis of this study is that countries with a lockdown decree can achieve greater success in reducing the mobility of the population, and the second hypothesis is that Argentina, Brazil and Colombia have regional mobility patterns. The first hypothesis was partially confirmed (considering 10 countries in South America), and the results obtained in the spatial analyzes confirmed the second hypothesis. In general, the observed data shows that less rigid lockdown or social distancing measures are necessary, however, they are not sufficient to achieve a significant reduction in the circulation of people during the pandemic.