摘要Background Pediatric heart failure(PHF)carries a high mortality burden,yet the prognostic value of admission blood pres-sure(BP)remains poorly de? ned,and evidence-based thresholds for risk strati? cation are lacking.Methods This retrospective,multicenter cohort study included 2545 children diagnosed with PHF upon admission across 30 Chinese centers(2013–2022).BP was standardized into age-,sex-,and height-speci? c percentiles.Associations with in-hospital mortality were analyzed using restricted cubic splines(RCS),machine learning covariate selection,and sensitivity analyses,strati? ed by sex,age,and etiology.Results Systolic BP(SBP)percentiles showed a U-shaped association with mortality,with the lowest risk at the 63.8th per-centile.Risk increased below the 57th percentile[adjusted hazard ratio(aHR)=1.86,95％con? dence interval(CI)1.07–3.26]and above the 91st percentile(aHR=1.69,95％CI=0.85–3.33),a relationship primarily driven by infants/toddlers.Diastolic BP(DBP)percentiles showed a linear inverse association,with risk increasing below the 83rd percentile(aHR=1.65,95％CI=1.09–2.50).Sex-strati? ed analyses suggested that girls were more vulnerable to high SBP(≥91st:aHR=3.84,95％CI=1.31–11.20),whereas boys were more susceptible to low DBP(<83rd:aHR=2.21,95％CI=1.16–4.20).Etiology-strati? ed analyses con? rmed consistency across major diagnostic categories(P values for all interactions>0.05).Conclusion This study establishes novel BP percentile thresholds for mortality risk strati? cation in patients with PHF(SBP:57th–91st;DBP:83rd–100th)and provides preliminary evidences for sex-and age-di? erential hemodynamic management,although validation is required.