In Kenya fluctuations of exchange rate caused economic policy debate involving citizens, policymakers, business community, academic researchers and business press with no empirical verification being done to authenticate their arguments. The study use ARDL analysis to analysis determinants of exchange rates in Kenya. Both ARDL long and error correction model were found to be robust because they passed all diagnostic tests such as serial correlation, heteroskedasticity and normality test. The CUSUM test confirmed the stability of both estimated models. The findings of this study addressed the existing knowledge gap in literature of on various determinants of exchange rate in Kenya. These will assist policy makers and government agencies, investors, financial institutions and multinational companies to know, estimate and predict determinants of foreign exchange fluctuations in Kenya.