We examine the long-run relationship between remittances and the real exchange rate for less-developed countries. In a key departure from the literature, we employ a panel cointegration approach using an innovative method for the measurement of the multilateral real effective exchange rate and we focus on high-remittance economies. We find a small inelastic, but significant, long-run relationship which confirms a Dutch disease type effect. The short-run relationship is explored using a panel vector error correction model which confirms that short-run causality is unidirectional running from remittances to the exchange rate. Potential asymmetries in this relationship are identified using quantile regression analysis.