OBJECTIVES: To compare the post-cessation weight gain following the use of different FDA-approved smoking cessation medication strategies among obese smokers.  METHODS: A retrospective cohort study was conducted using the General Electric (GE) electronic medical record database (2006 – 2011). The cohort consisted of obese adult smokers newly initiating use of an FDA-approved smoking cessation medication. The outcome variable was weight change at 3, 6, or 12 months following the first prescription. Descriptive analyses and t-tests were conducted to assess the frequency distribution of sample characteristics and their association with the post-cessation weight change. Multivariate linear regression models were carried out to identify predictors of weight change at 3, 6, and 12 months after assessing the model assumptions, with the use of multiple imputation to account for missing data for covariates.  RESULTS: The mean weight change was 1.14  (±17.26), 2.06 (±18.46), and 3.06 pounds (±20.78) at 3-, 6-, and 12-month, respectively. Obese smokers who were prescribed varenicline had a mean weight gain of 1.18 (±16.75), 2.14 (±18.14), and 3.12 pounds (±20.89) for each follow up, while those who were prescribed bupropion had a mean weight gain of 0.23 (±25.90), 0.22 (±25.32), and 1.47 pounds (±17.50), respectively. Descriptive analysis showed that obese smokers taking bupropion had less weight gain than those taking varenicline at each follow up; however, this association was not statistically significant after accounting for all covariates (β = -1.16 [-3.84 – -1.53] month 3; β = -3.16 [-6.54 – -0.21] month 6; β = -0.18 [-3.92 – 3.55] month 12). Significant predictors of weight change included: being diagnosed with diabetes, hyperlipidemia, taking weight-influencing medications, and smoked >= one cigarette/day. CONCLUSIONS: While patients using bupropion gained slightly less weight compared to those using varenicline, type of smoking cessation medication was not a significant predictor of weight change in the multivariate linear regression model.