BACKGROUND: Tofacitinib and inflammatory bowel disease (IBD) have been associated with increased risks for thromboembolic and cardiovascular events, but drug attributable risk is unknown. METHODS: We conducted a retrospective cohort study in a US claims database. We identified patients with IBD by International Classification of Disease (ICD) codes, stipulated 180 days of continuous enrollment prior to tofacitinib or anti-tumor necrosis factor (TNF) initiation to determine new users. Primary outcomes were ICD codes for venous thromboembolism (VTE) and cardiovascular (CV) events. We constructed propensity score (PS)-weighted Cox proportional hazard models to estimate hazard ratios (HRs) and time-to-event outcomes comparing tofacitinib and anti-TNF. We conducted a subgroup analysis of patients >/= 50 years. RESULTS: We identified 305 patients with IBD initiating tofacitinib and compared them with 19,096 initiating anti-TNFs. After weighting, balance was achieved across all demographic covariates. VTE occurred in 5% of patients treated with tofacitinib and 4% of anti-TNF users; in a PS-weighted cohort, tofacitinib did not confer a significantly elevated VTE risk compared with anti-TNF therapy (HR: 1.72, 95% CI: 0.74-3.01). A major CV event (MACE) occurred in 2% of tofacitinib users and 1% of anti-TNF users; tofacitinib also did not confer a significantly elevated risk for MACE (HR: 2.50, 95% CI: 0.37-6.18). Those with a Charlson comorbidity index >/= 2 had greater risks for thromboembolic and cardiovascular events. Similar findings were noted in patients >/= 50 years. CONCLUSIONS: In this large, active comparator, study, we demonstrate that tofacitinib was not associated with a higher risk of adverse thrombotic events compared with anti-TNFs in patients with IBD.