BACKGROUND: Hematoma expansion shift (HES) analysis can be used to assess the biological effect of a hemostatic therapy for intracerebral hemorrhage. In this study, we applied HES analysis to individual patient data from 4 randomized controlled trials evaluating rFVIIa (recombinant factor VIIa) 80 μg/kg to placebo. METHODS: We generated polychotomous strata of HES using absolute growth thresholds (≤0/<6/≥6 mL) and quintiles of percent volume change. The relationship between treatment and HES was assessed using proportional odds models. Differences in subgroups based on baseline volume (≥ or <20 mL), and time from symptom onset to treatment (≤ or >2 hours) were explored with testing for interactions. RESULTS: The primary analysis included 721 patients. At 24 hours, 36% (134/369) of rFVIIa-treated patients exhibited no hematoma expansion as compared with 25% of placebo (88/352)-treated patients. Significant expansion (≥6 mL) was reduced by 10% in those treated with rFVIIa-(adjusted common odds ratio [acOR], 0.57 [95% CI, 0.43–0.75]). An examination of percent change similarly showed a shift across the spectrum of expansion (acOR, 0.61 [95% CI, 0.47–0.80]). In both groups, mild-to-moderate expansion was observed in 38% to 47% of patients, depending on the threshold used. Differences in absolute HES between the rFVIIa and placebo groups were more pronounced in patients with baseline hemorrhage volumes ≥20 mL (acOR, 0.48 [95% CI, 0.30–0.76] versus <20 mL: acOR, 0.67 [95% CI, 0.47–0.95]; P _interaction =0.02). No treatment interaction in patients treated within 2 or after 2 hours from onset was observed (acOR, 0.42 [95% CI, 0.19–0.91 versus >2 hours: acOR, 0.59 [95% CI, 0.44–0.79]; P _interaction =0.30). CONCLUSIONS: The association between rFVIIa and hematoma growth arrest is most pronounced in patients with larger baseline volumes but is evident across the full spectrum of treated patients.