Context: Gastric bypass surgery is the most commonly performed bariatric surgical procedure in the United States. Variable weight loss following this relatively standardized intervention has been reported. To date, a method for reliable profiling of patients who will successfully sustain weight loss for the long term has not been established. In addition, the mechanisms of action in accomplishing major weight loss as well as the explanation for the variable weight loss have not been established. Objective: To examine whether gene expression in perioperative omental adipose is associated with gastric bypass-induced weight loss. Design: Cross-sectional study of gene expression in perisurgical omental adipose tissues taken/available at the time of operation and total excess weight loss (EWL). Subjects: Fifteen overweight individuals who underwent Roux-en-Y gastric bypass (RYGB) surgery at the University of California Davis Medical Center (BMI: 40.6-72.8 kg/m 2 ). Measurements: Body weight before and following weight stabilization 18-42 months after surgery. Perioperative omental adipose RNA isolated from 15 subjects was hybridized to Affymetrix HG-U133A chips for 22 283 transcript expression measurements. Results: Downstream analysis identified a set of genes whose expression was significantly correlated with RYGB-induced weight loss. The significant individual genes include acyl-coenzyme A oxidase 1 (ACOX1), phosphodiesterase 3A cGMP-inhibited (PDE3A) and protein kinase, AMP-activated, beta 1 non-catalytic subunit (PRKAB1). Specifically, ACOX1 plays a role in fatty acid metabolism. PDE3A is involved in purine metabolism and hormone-stimulated lipolysis. PRKAB1 is involved in adipocytokine signaling pathway. Gene network analysis revealed that pathways for glycerolipid metabolism, breast cancer and apoptosis were significantly correlated with long-term weight loss. Conclusion: This study demonstrates that RNA expression profiles from perioperative adipose tissue are associated with weight loss outcome following RYGB surgery. Our data suggest that EWL could be predicted from preoperative samples, which would allow for informed decisions about whether or not to proceed to surgery.