Antibiotic selection logic is grounded in matching the most likely pathogen(s) at a given infection site with an agent that achieves adequate tissue concentration while minimizing collateral damage and resistance development. Empiric therapy is chosen before culture results return and is guided by the most probable organisms for the clinical syndrome, local antibiogram data (institutional resistance patterns), infection severity (sepsis vs uncomplicated), and patient-specific factors such as allergies, renal/hepatic function, recent antibiotic exposure, and immunosuppression. Once culture and sensitivity data are available, the clinician de-escalates to the narrowest-spectrum agent with proven in-vitro activity—this targeted approach reduces Clostridioides difficile risk, limits selection pressure for multidrug-resistant organisms, and decreases cost. Pharmacokinetic-pharmacodynamic (PK/PD) optimization is essential: beta-lactams exhibit time-dependent killing (efficacy determined by time above MIC), whereas aminoglycosides and fluoroquinolones exhibit concentration-dependent killing (efficacy driven by peak-to-MIC ratio). Understanding these principles informs dosing intervals, extended infusions, and dose adjustments for obesity, renal impairment, or augmented renal clearance in critically ill patients.