In this paper, a new variant of Teaching-Learning based Optimization (TLBO), termed as Elitist Teaching-Learning Opposition based (ETLOBA) Algorithm has been proposed for numerical function optimization. The proposed method is empowered with two mechanisms to reach the accurate global optimum with less time complexity. One of them is elitism, which strengthens the capability of optimization method by retaining the best solution obtained so far, on the other hand Opposition method helps in ameliorating the capability of searching. As ETLOBA had an advantage of both Elitism and Opposition based learning, hence it tries to obtain optimum solutions with guaranteed convergence. The proposed method has been tested on several benchmark functions and the results obtained by ETLOBA are been compared with new state-of-art optimization methods like ABC, HS etc., shows the superiority of the proposed approach in solving continuous optimization problems.