Baseline characteristics of patients were similar between groups, but mean hematoma volumes were smaller in the guideline group (12.7 mL, SD 11.6) than in the intensive group (14.2 mL, SD 14.5). From randomisation to 1 h, mean systolic BP was 153 mm Hg in the intensive group and 167 mm Hg in the guideline group (difference 13.3 mm Hg, 95% CI 8.9-17.6 mm Hg; p<0.0001); from 1 h to 24 h, BP was 146 mm Hg in the intensive group and 157 mm Hg in the guideline group (10.8 mm Hg, 95% CI 7.7-13.9 mm Hg; p<0.0001). Mean proportional haematoma growth was 36.3% in the guideline group and 13.7% in the intensive group (difference 22.6%, 95% CI 0.6-44.5%; p=0.04) at 24 h. After adjustment for initial haematoma volume and time from onset to CT, median haematoma growth differed between the groups with p=0.06; the absolute difference in volume between groups was 1.7 mL (95% CI -0.5 to 3.9, p=0.13). Relative risk of haematoma growth >or=33% or >or=12.5 mL was 36% lower (95% CI 0-59%, p=0.05) in the intensive group than in the guideline group. The absolute risk reduction was 8% (95% CI -1.0 to 17%, p=0.05). Intensive BP-lowering treatment did not alter the risks of adverse events or secondary clinical outcomes at 90 days. Early intensive BP-lowering treatment is clinically feasible, well tolerated, and seems to reduce haematoma growth in ICH. A large randomised trial is needed to define the effects on clinical outcomes across a broad range of patients with ICH.