Recognition of gastric epithelial dysplasia, although a key to cancer prevention, can be challenging. In this study, we evaluated whether Lgl2 can serve as a marker of gastric foveolar-type dysplasia. Since atypical protein kinase C (aPKC) is a partner of Lgl2 in the control of apical-basal polarity we also investigated whether aPKC-zeta can compliment Lgl2 as a marker of dysplasia. Routinely processed specimens included 64 normal mucosa, 35 reactive gastropathies, 31 chronic gastritides, 65 gastric dysplasias (25 foveolar; 40 adenomatous), and 34 gastric adenocarcinomas. Twenty (80%) foveolar-type dysplasias showed absence of Lgl2 immunoreactivity, while normal basolateral expression of Lgl2 was consistently seen in normal gastric epithelium (n=20) and chronic gastritis (n=22; p<0.00001). Loss of Lgl2 was similar in the groups with low-grade and high-grade foveolar-type dysplasia, 79% and 83%, respectively. Linear apical anti-aPKC-zeta immunoreactivity was consistently present in the normal epithelium and was preserved in 91% of reactive gastropathies and 87% of chronic gastritides. In contrast, loss of apical aPKC-zeta staining was observed in 47% and 65% of low-grade dysplasias of foveolar and adenomatous types, respectively (p<0.005) and in nonsignificantly higher percentage of high-grade dysplasias. Apical aPKC-zeta staining was lost in 97% of gastric adenocarcinomas. Our data suggest a role of Lgl2 immunohistochemistry as an adjunct in the diagnosis of foveolar-type gastric dysplasia. aPKC-zeta had moderate sensitivity as a marker of gastric dysplasia and additional studies are needed to establish its role in the diagnosis of dysplasia.