Observational data show that nonsteroidal anti-inflammatory drug (NSAID) use is associated with a lower rate of breast cancer. We evaluated the effect of etodolac, an FDA-approved NSAID reported to inhibit cyclooxygenase (COX) enzymes and the retinoid X receptor alpha (RXR), on rationally identified potential biomarkers in breast cancer. Patients with resectable breast cancer planned for initial management with surgical resection were enrolled and took 400 mg of etodolac twice daily prior to surgery. Protein and gene expression levels for genes related to COX-2 and RXRα were evaluated in tumor samples from before and after etodolac exposure. Thirty subjects received etodolac and 17 subjects were assayed as contemporaneous or opportunistic controls. After etodolac exposure mean cyclin D1 protein levels, assayed by immunohistochemistry, decreased (P = 0.03). Notably, pre- versus post cyclin D1 gene expression change went from positive to negative with greater duration of etodolac exposure (r = -0.64, P = 0.01). Additionally, etodolac exposure was associated with a significant increase in COX-2 gene expression levels (fold change: 3.25 [95% CI: 1.9, 5.55]) and a trend toward increased β-catenin expression (fold change: 2.03 [95% CI: 0.93, 4.47]). In resectable breast cancer relatively brief exposure to the NSAID etodolac was associated with reduced cyclin D1 protein levels. Effect was also observed on cyclin D1 gene expression with decreasing levels with longer durations of drug exposure. Increased COX-2 gene expression was seen, possibly due to compensatory feedback. These data highlight the utility of even small clinical trials with access to biospecimens for pharmacodynamic studies.