The relationship between blood pressure variability (BPV) and stroke recurrence is controversial. This study aimed to evaluate the predictive value of short-term and long-term BPV for stroke recurrence. The data from blood pressure and clinical outcome in TIA or ischemic stroke (BOSS) study included participants with acute ischemic stroke or transient ischemic attack (TIA) within 7 days after onset. Short-term BPV was measured by 24-h ambulatory BP measurement during hospitalisation. Long-term BPV, based on home blood pressure measurement, was measured by patients or their families from discharge to 90 days after qualifying events onset (twice daily). Variability for systolic blood pressure (SBP) was assessed as the s.d., coefficients of variance (CV). The clinical outcome was recurrent stroke within 1 year. Cox proportional hazards models were used to test the association of short-term BPV, long-term BPV and stroke recurrence. Among 1764 participants, the mean age was 62.45 +/- 10.99 years, and 32.1% were females, during 1-year follow-up, 106 (6.0%) participants experienced recurrent stroke. Indices of long-term BPV were significantly associated with stroke recurrence (s.d.: adjusted HR: 1.939, 95% CI: 1.122-3.351; CV: adjusted HR 1.955, 95% CI: 1.174-3.255), independent of mean SBP. However, all indices of short-term BPV (s.d., CV) were not associated with stroke recurrence (s.d.: adjusted HR: 1.245, 95% CI: 0.680-2.280; CV: adjusted HR: 1.142, 95% CI: 0.640-2.038). In conclusion, long-term variation of SBP is a better predictor than short-term variation of SBP for recurrent stroke after acute ischemic stroke or TIA.