Solving linear diophantine equations in two variables have applications in computer science and mathematics. In this paper, we revisit an algorithm for solving linear diophantine equations in two variables, which we refer as DEA-R algorithm. The DEA-R algorithm always incurs equal or less number of recursions or recursive calls as compared to extended Euclidean algorithm. With the objective of taking advantage of the less number of recursive calls , we propose an optimized version of the DEA-R algorithm as DEA-OPTD. In the recursive function calls in DEA-OPTD, we propose a sequence of more efficient computations. We do a theoretical comparison of the execution times of DEA-OPTD algorithm and DEA-R algorithm to find any possible bound on the value of c for DEA-OPTD being better than DEA-R. We implement and compare an iterative version of DEA-OPTD (DEA-OPTDI) with two versions of a widely used algorithm on an specific input setting. In this comparison, we find out that our algorithm outperforms on the other algorithm against at least 96% of the inputs.