Mountain protected areas are popular tourism destinations. Those managing such landscapes and activities require data on when and where people go, but obtaining such information is difficult for large and often remote areas, particularly when resources are limited. We illustrate how geolocated images posted on social media could be used to assess tourism by analyzing geolocated images taken in the Annapurna Conservation Area in Nepal and posted to the image-sharing platform Flickr. Data from 577 visitors covering 3 major periods—pre-2007 (limited roads and Maoist insurgency), 2008–2014 (post-insurgency and better roads), and 2015–2020 (post-blizzard and earthquake)—and 4 meteorological seasons were analyzed. We found monthly patterns based on photo user days correlated with recent official visitor data. When finer resolution mapping of photo-user-hour data was visualized using heatmaps, seasonal and spatial patterns were apparent. Visitors mainly stayed on trails, roads, or in villages, with few images off trails, at high elevations, or from the remote north, while visitation was greater and more dispersed in spring and autumn compared to summer and winter. The results highlight the concentrated nature of visitation and hence opportunities to promote new destinations, events, and activities, particularly in summer and winter. They also show how roads and trails have expanded the use of the area. Although beneficial, social media has limitations, as few visitors post to social media, platforms vary in popularity and access to data, and there are increasing ethical and privacy issues with data from social media that need to be considered.