Background: Fatal bleeding is a serious consequence of anticoagulant therapy, but factors associated with fatal bleeding during the first 3 months of treatment of venous thromboembolism (VTE) are uncertain. Methods: Using data from RIETE, an ongoing registry of consecutive patients with acute VTE, we assessed risk factors for fatal bleeding among all patients. We then used this information to derive a clinical model that would stratify a patient’s risk of fatal bleeding during the first 3 months of treatment. Results: Of 24 395 patients, 546 (2.24%) had a major bleed and 135 (0.55%) had a fatal bleed. The gastrointestinal tract was the most common site (40% of fatal bleeds), followed by intracranial bleeding (25%). Fatal bleeding was independently associated with the following factors at the time of VTE diagnosis: age >75 years (OR, 2.16), metastatic cancer (OR, 3.80), immobility ≥ 4 days (OR, 1.99), a major bleed within the past 30 days (OR, 2.64), an abnormal prothrombin time (OR, 2.09), a platelet count < 100 × 109 L−1 (OR, 2.23), creatinine clearance < 30 mL min−1 (OR, 2.27), anemia (OR, 1.54), and distal deep vein thrombosis (OR, 0.39). INR at the time of bleeding is not known. A clinical prediction rule for risk of fatal bleeding that included nine baseline factors was derived. Fatal bleeding occurred in 0.16% (95% CI, 0.11–0.23) of the low‐risk, 1.06% (95% CI, 0.85–1.30) of the moderate‐risk, and 4.24% (95% CI, 2.76–6.27) of the high‐risk category. Conclusions: Patient characteristics and laboratory variables can identify patients at high risk for fatal bleeding during treatment of VTE.