We study the influence of the diaspora remittance flows on the intensity of conflict in the Sri Lankan Civil War during the period 2000-2009. Using data on Facebook connections at the subnational level, we infer which remittance flows were likely to benefit the Tamil Tiger rebels relative to the central government. Using shocks to source country remittance flows, we show that exogenous increases in remittances to Tamil Tiger-controlled areas significantly increased their fighting strength. We then set up a quantitative model of armed conflict between two sides and many contested geographical locations, augmented with remittance flows that affect the fighting strength of the two sides, and calibrate it to the Sri-Lankan Civil War. After structurally estimating key parameters using remittance and fighting outcome data, we use the model to quantify how war outcomes would have differed in the absence of remittance flows. We find that remittances had a significant impact on the timing of the central government victory. Remittances that favoured the Tamil Tiger rebels prolonged the war by as much as 9 years.