The objective of this study was to determine the pharmacokinetic profile of meropenem in automated peritoneal dialysis (APD) patients. In 6 patients without peritonitis, a single dose of 0.5 g of meropenem was applied intraperitoneally (i.p.) or intravenously (i.v.) and concentrations in serum and dialysate were measured at specified intervals over 24 h with high-performance liquid chromatography-mass spectrometry. The mean maximum concentrations of meropenem in serum (Cmax) were 27.2 mg/liter (standard deviation [SD], ±6.9) and 10.1 mg/liter (SD, ±2.5) and in dialysate were 3.6 mg/liter (SD, ±2.3) and 185.8 mg/liter (SD, ±18.7) after i.v. and i.p. administrations, respectively. The mean areas under the curve from 0 to 24 (AUC0-24) of meropenem in serum were 173.5 mg · h/liter (SD, ±29.7) and 141.4 mg · h/liter (SD, ±37.5) (P = 0.046) and in dialysate were 42.6 mg · h/liter (SD, ±20.0) and 623.4 mg · h/liter (SD, ±84.1) (P = 0.028) after i.v. and i.p. administrations, respectively. The ratios for dialysate exposure over plasma exposure after i.v. and i.p. treatments were 0.2 (SD, ±0.1) and 4.6 (SD, ±0.9), respectively (P = 0.031). A mean target value of 40% T>MIC (time for which the free meropenem concentration exceeds the MIC) for clinically relevant pathogens with EUCAST susceptibility breakpoints of 2 mg/liter was reached in serum after i.p. and i.v. administrations and in dialysate after i.p. but not after i.v. administration. The present data indicate that low i.p. exposure limits the i.v. use of meropenem for PD-associated peritonitis. In contrast, i.p. administration not only results in superior concentrations in dialysate but also might be used to treat systemic infections.