Starting in 2011, the Syrian civil war has resulted in the displacement of over 80% of the Syrian population. This paper analyzes how the widespread use of social media has recorded migration considerations for Syrian refugees using social media text and image data from three popular platforms (Twitter, Telegram, and Facebook). Leveraging survey data as a source of ground truth on the presence of IDPs and returnees, it uses topic modeling and image analysis to find that areas without return have a higher prevalence of violence-related discourse and images while areas with return feature content related to services and the economy. Building on these findings, the paper first uses mixed effects models to show that these results hold pre- and post- return as well as when migration is quantified as monthly population flows. Second, it leverages mediation analysis to find that discussion on social media mediates the relationship between violence and return in months where there are fewer violent events. Monitoring refugee return in war prone areas is a complex task and social media may provide researchers, aid groups, and policymakers with tools for assessing return in areas where survey or other data is unavailable or difficult to obtain.