Abstract Objectives We aimed to identify rational empirical dosing strategies for cefepime treatment in critically ill patients by utilizing population pharmacokinetics and target attainment analysis. Patients and methods A prospective and opportunistic pharmacokinetic (PK) study was conducted in 130 critically ill patients in two ICU sites. The plasma concentrations of cefepime were determined using a validated LC-MS/MS method. All cefepime PK data were analysed simultaneously using the non-linear mixed-effects modelling approach. Monte Carlo simulations were performed to evaluate the PTA of cefepime at different MIC values following different dose regimens in subjects with different renal functions. Results The PK of cefepime in critically ill patients was best characterized by a two-compartment model with zero-order input and first-order elimination. Creatinine clearance and body weight were identified to be significant covariates. Our simulation results showed that prolonged 3 h infusion does not provide significant improvement on target attainment compared with the traditional intermittent 0.5 h infusion. In contrast, for a given daily dose continuous infusion provided much higher breakpoint coverage than either 0.5 h or 3 h intermittent infusions. To balance the target attainment and potential neurotoxicity, cefepime 3 g/day continuous infusion appears to be a better dosing regimen than 6 g/day continuous infusion. Conclusions Continuous infusion may represent a promising strategy for cefepime treatment in critically ill patients. With the availability of institution- and/or unit-specific cefepime susceptibility patterns as well as individual patients’ renal function, our PTA results may represent useful references for physicians to make dosing decisions.