DescriptionLord’s Wald test for differential item functioning (DIF) has not been extensively studied particularly in the context of multidimensional IRT (MIRT) framework. Lord’s Wald test was implemented using two estimation approaches in the MIRT framework: Marginal maximum likelihood (MML) estimation based on expectation maximization (EM) algorithm and the Bayesian Markov chain Monte Carlo (MCMC) estimation based on Metropolis-Hastings algorithm. This study investigated the recovery of item parameters, the Type I error, and the power of Lord’s Wald tests obtained from the two estimation approaches under various simulation conditions, including DIF type differences, DIF magnitude differences, test length differences, and different combinations of sample sizes. Item responses were generated under multidimensional two-parameter logistic and three-parameter logistic models. Specific concerns for designing DIF detection conditions in MIRT framework were outlined based on the literature review on unidimensional and multidimensional DIF methods. The relative performances of the two estimation methods compared and summarized under the simulation conditions considered in this study. Furthermore, English usage data were used to illustrate the use of Lord’s Wald test with the two estimation approaches. Finally, the summary and implications of the results, the limitations of the present study, and directions for further studies were discussed.