Bloodstream infections (BSIs) are life-threatening infections for which a timely initiation of appropriate antimicrobial therapy is critical. Antibiotic susceptibility testing (AST) directly performed on positive blood culture broths can help initiate targeted antibiotic therapy sooner than the standard AST performed on colonies isolated on solid media after overnight incubation. Faster antimicrobial susceptibility testing (AST) results can improve clinical outcomes, and reduce broad-spectrum antimicrobial consumption and healthcare-associated costs in sepsis. In this study, we evaluated the accuracy of a direct AST inoculation method on the BD Phoenix M50 system using serum separator tubes to harvest bacteria from positive pediatric blood culture bottles. Direct AST was performed on 132 monomicrobial pediatric blood culture bottles that were positive for Enterobacterales (65; 49.2%), Staphylococcus aureus (46; 34.8%), and non-fermenting Gram-negative bacilli (21; 16%). Overall, the categorical and essential agreements between the direct method and standard method were 99.6% and 99.8%, respectively. Very major, major, and minor error rates were 0.1%, 0.09%, and 0.20% respectively. Direct AST performed on pediatric blood culture bottles using BD Phoenix M50 can quickly provide accurate susceptibility information to guide antimicrobial therapy in patients with BSI.