Alterations in the mismatch repair genes (hMLH1 and hMSH2) play an important role in the development of microsatellite instability in sporadic endometrial cancer. Tissue microarray technology allows molecular profiling of tumor samples at the DNA, RNA, and protein levels. We analyzed hMLH1 and hMSH2 expression by immunohistochemistry in a group of atypical endometrial hyperplasias (n = 10), endometrioid endometrial carcinomas (n = 58), and nonendometrioid endometrial carcinomas (n = 27) on tissue microarray. The results were correlated with microsatellite instability status as evaluated by BAT-25 and BAT-26. Overall, 29.4% of lesions showed microsatellite instability. Loss of nuclear hMLH1 and hMSH2 protein expression was seen in 22.3% and 6.5% of cases, respectively. Immunohistochemistry for hMLH1 and hMSH2 showed lack of protein expression in 64% and 16.6% of microsatellite instability-positive endometrial lesions, respectively. Taken together, hMLH1 or hMSH2 protein expression was absent in 18 of 24 microsatellite instability-positive cases (75% sensitivity). A high level of concordance was found between immunohistochemistry for hMLH1 and hMSH2 and microsatellite instability status evaluated by BAT-25 and BAT-26 (kappa value of 0.7). Of the 57 cases found to be microsatellite instability negative, 53 showed normal expression of both proteins (93% specificity). The observed predictive value of absence of expression of hMLH1 for predicting microsatellite instability-positive status was 82%. The predictive value of normal expression of both proteins for predicting microsatellite instability-negative status was 90%. These results are consistent with those previously reported in whole tissue sections. Therefore, immunohistochemical analysis of hMLH1 and hMSH2 expression on tissue microarray provides an accurate technique for screening for tumors with microsatellite instability. Tissue microarrays represent an ideal approach for comparing different diagnostic or predictive markers with one another in consecutive tissue microarray sections.