In this work, we advance the solution approach to the single stage capacitated warehouse location problem (SSCWLP) problem. We took six formulations given by Sharma and Berry (2007) for SSCWLP with various constraints combinations and added a new promising constraint (∑capj * yj  ≥  1). The computations done after additional constraint results in statistically better results reduction in number of nodes searched as well as in CPU time taken to solve the problem. We created random 50 samples of problems of size 15x15x15, 30x30x30 and 50x50x50 size problems and solved using GAMS software to show the computational advantage even in such large magnitude problem. We performed paired t-test on all the six pairs of results to observe statistical significance of improvement with the new constraint. Results with new constraint were observed statistically significant and better at two-tailed t-test with 95% confidence interval.