Aims: Concerning the CEMAC area in Central Africa, about the recovering of parameters estimates in the Long Run Relationship of Exports equation, the conventional cointegration techniques have been used as well as in time series than panel data framework. Nevertheless recent developments in econometrics reveal that this conventional techniques particularly in times series suffers from the issue to be more sensitive to finite sample biais. Then the aim of this paper is to revisist the Long Run Relationship of Exports equation at the light of this new developments of econometric techniques for assessing long run relationships. Method: Then the literature now prefers to use the well known approach of ARDL modelling. In fact, this is presented as robust to small sample biais. Then this Modelling have been extendend to a panel data framework with the Pooled Mean Group (PMG) and Mean Group (MG) estimates Main findings: The data spams the period 1974-2021 for the PMG and the MG estimates and 2009-2019 for the tests of existence of long run relationships among the variables. Globally speaking, from the results obtained here the assumption of unitary foreign demand elasticity is well admitted, the price competitiveness channel holds and the classical result of substituability between domestic demand and exports is observed contrarely to the liquidity constraint assumption. Conclusion: The ARDL Modelling is then proove to be an flexible method than can be apply to a wide variety of economic problems as those presented here about the Long Run Relationship in Export Equation as well as in a times series framework than panel data modelling.