Abstract Background Among patients undergoing resection for gastric cancer, the impact of body mass index (BMI) on outcomes is not well understood. We sought to define the impact of non-normal BMI on short- and long-term outcomes after gastric cancer resection. Methods We identified 775 patients who underwent gastrectomy for adenocarcinoma between 2000 and 2012 from the multi-institutional US Gastric Cancer Collaborative. Clinicopathologic characteristics, operative details, and oncologic outcomes were collected, and patients were stratified according to BMI. Results Most patients in the cohort were classified as having normal BMI ( n  = 338, 43.6%), followed by overweight ( n  = 229, 29.6%), obese ( n  = 153, 19.7%), and underweight ( n  = 55, 7.1%). After stratifying by BMI, there were no significant differences in the incidence of postoperative blood transfusions, perioperative morbidity, postoperative infectious complications, length of stay, perioperative 30-d in-hospital death, or readmission across groups (all P  > 0.05). BMI did not impact overall or recurrence-free survival after stratifying by stage (all P  > 0.05). However, underweight patients with low preoperative albumin levels had worse overall survival (OS) compared with that of patients of normal BMI. Conclusions BMI did not impact perioperative morbidity, recurrence-free, or OS in patients undergoing gastric resection for adenocarcinoma. Underweight patients with BMI <18.5 kg/m2 and low preoperative albumin levels, however, had a significantly decreased OS after gastrectomy for cancer. These high-risk patients should have their nutritional status optimized both before and after gastrectomy in an attempt to modify this risk factor and, in turn, achieve better outcomes.