Most patients with gastric cancer rapidly lose weight after gastrectomy. Therefore, analysis of the effect of body mass index (BMI) on patients with gastric cancer survival should include postoperative BMI and BMI loss and preoperative BMI. This retrospective cohort study analyzed the effect of three BMI variables and their interaction on long-term outcomes. Preoperative BMI analysis included 2,063 patients with gastric cancer who underwent curative gastrectomy between January 2009 and December 2013 at Seoul National University Bundang Hospital. BMI at postoperative 6 to 12 months was available in 1,845 of these cases. Patients with preoperative BMI 23.0 to <27.5 HR, 0.63; 95% confidence interval (CI), 0.48-0.82 for BMI 23.0 to <25.0 and HR, 0.57; 95% CI, 0.42-0.78 for BMI 25.0 to <27.5 and postoperative BMI 23.0 to <25.0 (HR, 0.67; 95% CI, 0.46-0.98) showed significantly better overall survival (OS) than pre- and postoperative patients with BMI 18.5 to <23.0, respectively. Postoperative underweight (BMI <18.5; HR, 1.74; 95% CI, 1.27-2.37) and postoperative severe BMI loss (>4.5; HR, 1.79; 95% CI, 1.29-2.50) were associated with higher mortality. Severe BMI loss and preoperative BMI <23.0 had an adverse synergistic effect; patients with BMI <23.0 were more vulnerable to severe BMI loss than those with BMI ≥23.0. Associations with cancer-specific survival were similar. All three BMI variables were prognostic factors for survival of patients with gastric cancer. Preoperative BMI and severe BMI loss had an interaction. Perioperative BMI and weight loss should be analyzed collectively in patients with gastric cancer undergoing gastrectomy. .