Objective: Although convincing evidence on blood pressure (BP) targets in older adults is increasing, data of individual BP management in this group is lacking, especially in frail subjects. We aimed to investigate whether frailty modifies the association between systolic blood pressure (SBP) and all-cause mortality in community-dwelling older adults. Design and method: This study included participants aged 65 years or older from the Chinese Longitudinal Healthy Longevity Survey 2002–2014 and followed up to 2018. They were divided into two groups according to a frailty index based on the accumulation of 44-items deficits. The association between SBP and mortality was analyzed using Cox proportional hazards models adjusted for age, sex, marital status, education, residence, economic income, current smoking, current alcohol consumption, BMI, DBP, cognitive impairment, hypertension, diabetes, cardiovascular disease, stroke and other cerebrovascular diseases, respiratory disease, and cancer. Results: Among 18,503 participants included, 75.2% of them were 80y or older, and 42.2% were frail. The median follow-up time was 42.7 months, detecting 15,901 (85.9%) deaths. We identified 7808 (42.2%, 7533 deaths) frail participants (mean frailty index = 0.33). Effect modification by frailty was detected. Among frail participants, a U-shaped association was found with hazard ratios of 1.16 (95% CI, 1.02–1.32) for SBP < 100 mmHg, and 1.11 (95% CI, 1.00–1.24) for SBP 150 mmHg or higher compared with SBP 120–130 mmHg. For non-frail older adults, a tendency toward higher risk among those with SBP 130 mmHg or higher was observed. Conclusions: Our results suggest the presence of effect modification by frailty indicating a possible negative effect for elevated SBP in non-frail older adults and a U-shaped relationship of SBP in frail older adults with respect to all-cause mortality even after adjusting for diastolic blood pressure.