<b>Objectives:</b> Endometrial intraepithelial neoplasia (EIN), including simple hyperplasia and complex hyperplasia with atypia (CAH), is a precursor lesion for endometrial adenocarcinoma. Not all patients with EIN are eligible for or choose to have a hysterectomy and are typically treated with progestin. Depending on EIN classification and covariates like obesity, 25-50% of cases fail to respond to progestin therapy. We hypothesized that non-responders might have mutations in <i>PTEN</i> that affect function but not expression. <b>Methods:</b> This is a retrospective case-control study of 49 patients who had <i>de novo</i> endometrial biopsies with EIN (41 CAH and eight simple hyperplasia). Included cases required diagnoses confirmed by two pathologists and a history of oral or IUD progestin therapy. All the cases had at least 3-6 months of follow-up biopsies. Regions of interest containing EIN were sampled for DNA analysis using histologic sections and the commercially available GeneTrails<sup>c</sup> solid tumor panel. Protein expression was measured in serial sections by Nanostring immunolabeling and GeoMx spatial profiling. Data were analyzed by Chi-square and unsupervised clustering with adjusted p-values. <b>Results:</b> Clinical follow-up revealed that 24 patients were progestin responders (eight simple hyperplasia and 16 CAH), and 25 were progestin non-responders (all with CAH). Those who did not respond to progestin either progressed to adenocarcinoma by three months or had persistent hyperplasia by six months. <i>PTEN</i> mutations above the variable allele frequency were identified in 19/25 of the progestin non-responders and 2/16 of the progestin responders yielding a likelihood ratio of 6 (2-23, p<0.001). Nanostring immunolabeling revealed no difference in <i>PTEN</i> expression between groups. <b>Conclusions:</b> Although there are no significant differences in <i>PTEN</i> expression in progestin non-responders compared with progestin responders for patients with EIN, there is a significantly increased frequency of <i>PTEN</i> functional mutations in the non-responders. This data suggests that genomic information, specifically <i>PTEN</i> mutations, predicts treatment response to progestin in EIN and has the potential to guide treatment decisions.