Here we describe the pharmacokinetics of piperacillin administered by continuous infusion (CI) versus intermittent bolus (IB) dosing in critically ill patients receiving continuous venovenous haemofiltration (CVVH) and compare the frequency of pharmacodynamic/pharmacokinetic (PK/PD) target attainment with each dosing strategy. This was a prospective pharmacokinetic trial in 16 critically ill patients with severe sepsis or septic shock undergoing CVVH and randomised to receive either CI or IB administration of a standard daily dose of piperacillin/tazobactam (11.25g/day on Day 1 followed by 9g/day). Serial blood samples were measured on two occasions. Piperacillin pharmacokinetics were calculated using a non-compartmental approach. Blood concentrations were compared with established PK/PD targets. On occasion 1 (Days 1-3 of therapy), IB administration resulted in significantly higher piperacillin peak concentrations (169 vs. 89mg/L; P=0.002), whereas significantly higher steady-state concentrations were observed in CI patients (83 vs. 57mg/L; P=0.04). Total clearance and clearance not mediated by CVVH were significantly higher with CI administration [median (interquartile range), 1.0 (0.7-1.1) and 0.8 (0.6-1.0)mL/kg/min; P=0.001 and 0.001, respectively]. The estimated unbound piperacillin concentrations were four times above the target susceptibility breakpoint (16mg/L) for the entire dosing interval (100%fT>4xMIC) in 87.5% of patients receiving CI administration (sampling occasion 1), compared with 62.5% of IB patients achieving the desired target (50%fT>4xMIC). Compared with IB dosing, and despite similar CVVH settings, CI administration of piperacillin results in a pharmacokinetic profile that may optimise outcomes for less susceptible pathogens.