A simulation study on two sample robust permutation tests for location is presented. The robust statistics studied fall into three categories; L-estimators, M-estimators, and adaptive estimators. The normal, Laplace, Cauchy, and mixture of normal distributions are used to investigate the power of the robust permutation tests, and normal distributions contaminated with outliers are used to investigate the effect of outliers on type-I error rates. The Wilcoxon rank sum test is used as a standard against which to measure the performance of the robust permutation tests