Existing panel data studies of real interest parity are either unable to identify which panel members are characterised by stationary real interest differentials, or are subject to size distortion resulting from the presence of structural breaks and cross-sectional dependencies. Using a panel stationarity testing procedure recently advocated by Hadri and Rao (2008) that allows for structural breaks and cross-sectional dependency, we are unable to reject the stationarity of Asian real interest rate differentials.