We aimed to assess the association between systolic blood pressure (SBP) and clinical outcome in 2,397 ischemic stroke (IS) or transient ischemic attack (TIA) patients from the Blood Pressure and Clinical Outcome in TIA or Ischemic Stroke (BOSS) study. BOSS study was a hospital-based, prospective cohort study. The SBP was defined as mean value of 90 days self-measured SBP after onset. Cox proportional hazards models were conducted to test the risk of combined vascular events (CVE) and stroke recurrence among different SBP categories. Restricted cubic splines were used to explore the shape of associations between SBP and clinical outcomes. A J-shaped association of SBP with CVE and stroke recurrence within 90 days was observed (P nonlinearity < 0.001 for both). After adjusting for age, gender, medical history, atrial fibrillation, admission NHISS score, and secondary prevention. The hazard ratios (95% confidence intervals) of SBP <115 and ⩾165 mmHg compared with 125-134 mmHg were 3.45 (1.11-10.66) and 7.20 (2.91-17.80) for CVE, 2.68 (0.75-9.53) and 9.69 (3.86-24.35) for stroke recurrence, respectively. Similar J-shaped relationships were found after 1 year of follow-up. In conclusion, both high and low SBP are associated with poor prognosis in this population.