This article is timely to evaluate Active Labor Market Policies (ALMPs) in Tunisia in 7 regions throughout the period from 2005 to 2014 with reference to OECD countries. Massive unemployment in Tunisia resisted during this period (15,3% in 2014), followed by major changes after the transition period. In order to study the efficiency levels achieved by the Tunisian Active Labor Market Policies, we propose a parametric method, namely the Stochastic Frontier Approach 'SFA'. Our results show that the total ALMPs significantly affects the unemployment rate. Thus, the average efficiency score of ALMP in Tunisia (0,94) is lower than that of the OECD countries (0,97). The divergence is due to structural and cyclical differences, which highlights the imperative to adopt a model adequate to the specificity of Tunisia. Tunisian public employment service policies have been shown to be insignificant compared to those of OECD countries. Start-up incentives seem to be the most effective policy to reduce the unemployment rate in the OECD countries' and Tunisia because of its low cost. [ABSTRACT FROM AUTHOR]