Aims This study aimed to characterize the population pharmacokinetic parameters of intravenously administered amikacin in newborns and assess the effect of sepsis in amikacin exposure. Methods Newborns aged ≥3 days who received at least 1 dose of amikacin during their hospitalization period were eligible for the study. Amikacin was administered intravenously during a 60‐min infusion period. Three venous blood samples were taken from each patient during the first 48 h. Population pharmacokinetic parameter estimates were obtained using a population approach with the programme NONMEM. Results Data from 329 drug assay samples were obtained from 116 newborn patients (postmenstrual age [PMA] 38.3, range 32–42.4 weeks; weight 2.8, range 1.6–3.8 kg). Measured amikacin concentrations ranged from 0.8 to 56.4 mg/L. A 2‐compartment model with linear elimination produced a good fit of the data. Estimated parameters for a typical subject (2.8 kg, 38.3 weeks) were clearance (Cl = 0.16 L/h), intercompartmental clearance (Q = 0.15 L/h), volume of distribution of the central compartment (Vc = 0.98 L) and peripheral volume of distribution (Vp = 1.23 L). Total bodyweight, PMA and the presence of sepsis positively influenced Cl. Plasma creatinine concentration and circulatory instability (shock) negatively influenced Cl. Conclusion Our main results confirm previous findings showing that weight, PMA and renal function are relevant factors influencing newborn amikacin pharmacokinetics. In addition, current results showed that pathophysiological states of critically ill neonates, such as sepsis and shock, were associated with opposite effects in amikacin clearance and should be considered in dose adjustments.