The efficacy of pembrolizumab in patients with microsatellite instability (MSI)-high cancers has been reported; however, the differences in efficacy according to the subtypes of MSI-high endometrial cancers (ECs) remain unclear. MSI-high ECs are classified into at least 3 groups based on their molecular characteristics: MLH1 hypermethylated, Lynch-like syndrome (LLS)-associated, and Lynch syndrome (LS)-associated cancers. This study aimed to investigate whether the efficacy of pembrolizumab differs among these 3 groups, and if so, whether EPM2AIP1 immunohistochemistry (IHC), which correlates with MLH1 promoter methylation, can be used to rule out MLH1 methylation cases. This study included 12 patients with MSI-high EC who received pembrolizumab treatment. Patients were categorized into 3 groups based on MLH1 methylation analysis and the Amsterdam Criteria: MLH1 hypermethylated (sporadic [SP]), LLS-associated, and LS-associated. Patients' medical records were retrospectively reviewed, and the efficacy of treatment was evaluated based on the response rate using the Response Evaluation Criteria in Solid Tumors version 1.1. The overall response rate was 75% (3/4) in the SP group, while it was 100% including one complete response patient in the LLS-associated and the LS-associated group, respectively. The sensitivity and positive predictive value of EPM2AIP1 IHC for MLH1 methylation were 100% and 66.7%, respectively. Pembrolizumab may be more effective in LLS and LS-associated groups. EPM2AIP1 IHC was less predictive than MLH1 methylation analysis; however, it may be useful for ruling out MLH1 methylation cases due to its high sensitivity. Further studies are needed to determine whether EPM2AIP1 IHC can predict pembrolizumab efficacy.