Delayed antimicrobial therapy is associated with poor outcomes in sepsis, but the optimal antibiotic administration time remains unclear. We aimed to investigate the effects of the time of antimicrobial administration on outcomes and evaluate an optimal empirical antibiotic administration time window for children with Streptococcus pneumoniae sepsis. This retrospective study enrolled children with S. pneumoniae sepsis who presented to the Children’s Hospital of Chongqing Medical University from May 2011 to December 2018. Classification and regression tree (CART) analysis was used to determine the time-to-appropriate-therapy (TTAT) breakpoint. Outcomes were compared between patients receiving early or delayed therapy, defined by CART-derived TTAT breakpoint. During the study period, 172 patients were included. The CART-derived TTAT breakpoint was 13.6 h. After adjustment for confounding factors, a TTAT of ≥13.6 hours was found to be an independent predictor of sepsis-related in-hospital mortality (odds ratio [OR] = 39.26; 95% confidence interval [CI] = 6.10 to 252.60), septic shock (OR = 4.58; 95% CI = 1.89 to 11.14), and requiring mechanical ventilation (OR = 2.70; 95% CI = 1.01 to 7.35). A Pediatric Risk of Mortality (PRISM) III score of ≥10 was independently associated with delayed therapy. Delayed antibiotic therapy was associated with poor outcomes in children with S. pneumoniae sepsis. The optimal empirical antibiotic administration time window in children with S. pneumoniae sepsis was within 13.6 h. Efforts should be made to ensure timely and appropriate therapy.