Objective To assess whether hemorrhage after early anticoagulation in nonseptic embolic infarction is related to clinical severity and infarction size. Background Explicit clinical criteria and timing of anticoagulation after large embolic infarctions are unknown. Methods Out of 171 patients receiving anticoagulation between July 1992 and December 1993, 83 patients with hemispheric embolisms received heparin within 72 hours from onset (activated partial thromboplastin time [aPTT] 1.5 times control value). Stratified by age and sex, a “high-risk” group (46 patients) was defined as those having stroke symptoms involving three or more CNS domains, Mathew Scale score ≤74, or hemorrhagic infarction on initial CT, and a “low-risk” group (37 patients) as those having stroke symptoms involving fewer than three cortical domains, or Mathew Scale score >74, and CT showing no blood. Loss of consciousness, seizures, or history of bleeding were exclusion criteria. Repeated CTs (100%) and MRIs (36%) detailed infarctions according to standard maps and evaluated all unexplained clinical worsening. Results Prior to therapy, high-risk patients had more severe clinical deficits (p p < 0.01), and more mass effect (p p < 0.01). Conclusions Infarction size and clinical severity in alert patients with nonseptic embolic stroke carries no additional bleeding complications after early anticoagulation. If anticoagulants are deemed necessary, treatment delay seems unjustified. However, rigorous monitoring of the aPTT is strongly advised to keep the level at 1.5 to 2 times control values.