Empirical studies report that there is a negative relationship between the spot difference and forward premium. This result violates the forward rate unbiasedness theory. Using standard regression we found that recent samples give mixed results with both positive and negative coefficients. One possibility is that the negative coefficients could arise due to the non-linearities in the series and misspecification. To overcome these problems we employed a relatively novel technique. As an alternative to the standard regression we used a time-varying coefficient technique. This methodology estimates bias-free coefficients and thus should provide better estimates of the link between spot and forward rates. The findings of the time-varying coefficient model strongly support the forward rate unbiasedness hypothesis. All the parameters are very close to unity and significant. At the same time our results do not violate the efficient market theory.