We provide a pioneer study to offer a solution to the popular UIP failure using risk premium in the component GARCH-in-mean framework. Using a dataset of 22 advanced and 22 emerging currencies, we report that in the absence of risk premium, UIP failure is more prominent in emerging countries relative to advanced countries and that 34% of the total beta coefficients range between 0.5 and 1.5. Next, we include the risk premium in the main UIP equation using a component GARCH-M model which allows decomposing the risk into permanent and transitory risks. The results become more consistent with the UIP theory and now almost 73% of the beta coefficients lie between 0.5 and 1.5. Such a finding validates our argument that risk premium is the main factor responsible for UIP violation and including it in the main equation helps uncover the UIP puzzle, especially in the case of emerging countries. This is our key contribution to the literature.