In Kenya, there are few empirical studies that focus on the relationship between budget and trade deficit. The existing empirical studies used panel and cross-section data to examine the relationship between budget deficit and trade deficit. The findings of these studies may not adequately capture the relationship between the two deficits since these countries have different levels of budget and trade deficits, different levels of economic development. In addition, the previous studies have not adequately addressed the problem of endogeneity. Further the findings of the previous results present mixed results. Some studies report unidirectional relationship and others bidirectional relationships between the budget deficit and trade deficit. The objective of this study therefore was to; determine causality between budget deficit and trade deficit in Kenya, determine the short-run and long-run effect between the twin deficits in Kenya and determine the speed of adjustment of budget and trade deficits in Kenya. The current proposed study, therefore, used annual time series data for the period 1970 to 2014 available in World Bank world development indicators CD-ROM, 2014. The data was analyzed first by testing for unit root using Phillip- Perron (PP) test approach. The variables were found to be integrated and were of the same order, and then test for the existence of co integration vector using the bounds test approach to co integration test technique was conducted. The variables were found to be co integrated and then error correction model was estimated using autoregressive distributed lag (ARDL) technique to address the problem of endogeneity. The analysis reveals that there is a positive short-run and long-run relationship between the trade deficit and the budget deficit. Further, the results show that there is bi-directional relationship between trade deficit and budget deficit. Budget deficit adjustment not trade deficit adjustment is shown to be the key engine governing the speed of budget-trade deficit convergence; that is, the budget deficit is the primary variable that changes in order to restore equilibrium when the system has been subjected to shock. Moreover, budget deficits are found to converge much faster than trade deficits.