Existing research exploring predictors of success on American Board of Surgery (ABS) exams focused on either resident or residency program characteristics, but limited studies focus on both. This study examines relationships between both resident and program characteristics and ABS Qualifying (QE) and Certifying Exam (CE) outcomes. Multilevel logistic regression was used to analyze the relationship between resident and program characteristics and ABS QE and CE 1st attempt pass and eventual certification. Resident characteristics were gender, IMG status, and prior performance, measured by 1st attempt USMLE Step 2 CK and Step 3 scaled scores. Program characteristics were size, %female, %International Medical Graduate (IMG), and program type. The sample included surgeons with QE and CE data from 2007-2019 and matched USMLE scores. Controlling for other variables, prior medical performance positively related to all ABS exam outcomes. The relationships between USMLE scores and success on ABS exams varied but were generally strong. Other resident characteristics that predicted ABS exam outcomes were gender and IMG (QE 1st attempt pass). The only program characteristic that significantly predicted ABS outcomes was %IMG (QE and CE 1st attempt pass). Despite statistical significance, gender, IMG, and %IMG translated to small differences in predicted probabilities of ABS exam success. This study highlights resident and program characteristics that predict success on ABS exams. USMLE scores consistently and strongly related to ABS exam success, providing evidence that USMLE scores relate to future high-stakes consequences like board certification. After controlling for prior performance, gender, IMG, and program %IMG significantly related to ABS exam success, but effects were small.