This thesis aimed at investigating the pharmacokinetics (PK) of antiinfectives at the site of infection. This objective was pursued in three different projects. In the first project, vancomycin and linezolid were studied in in vitro experiments to enable the conduction of clinical trials investigating their PK by using the microdialysis approach. A rapid and reliable HPLC assay capable of measuring vancomycin concentrations in microdialysate and plasma was developed. Moreover, a previously existing analytical assay for linezolid was extended to the matrices urine, bone marrow, bone biopsy samples and bone microdialysate. Microdialysis investigations were able to show that vancomycin is suitable for microdialysis experiments on the condition that Ringer’s solution is replaced by phosphate buffer in microdialysis perfusate. In the second project, microdialysis was applied to corticancellous bone tissue of healthy sows after single intravenous linezolid infusion. The in vivo study was able to demonstrate the feasibility and validity of the microdialysis technique in bone tissue. PK investigations revealed that linezolid did not penetrate into bone tissue to the extent that might have been expected from measuring plasma and homogenated tissue samples. AUC and Cmax values stayed considerably below those of all other matrices. By relating the results to pharmacokinetic indices such as AUC/MIC it was concluded that the standard linezolid dose might not be sufficient for the treatment of bone infections in both animals and humans. Finally, in the third project a clinical trial was conducted in order to assess the PK of linezolid in healthy volunteers and septic patients after single and multiple dosing. Unbound linezolid concentrations were determined in plasma as well as the interstitium (ISF) of subcutaneous adipose tissue and skeletal muscle applying the microdialysis technique. Using the population PK approach unbound linezolid plasma concentrations were characterised by a two-compartment model. The observed PK nonlinearity was attributed to a change in clearance, which presumably might be due to an inhibition of the respiratory chain enzyme activity in the course of linezolid treatment. It was accounted for by introducing an empirical inhibition compartment. ISF concentrations were implemented by the use of two additional compartments that were coded using monodirectional rate constants and partition coefficients. Overall, linezolid displayed excellect penetration abilities into both subcutaneous and muscular ISF. However, large variability was observed. Creatinine clearance, body weight and thrombocytes were able to explain some of the observed variability in clearance, peripheral volume of distribution, rate into the muscular compartment and on the partition coefficient into muscular ISF. These relations should be confirmed in subsequent trials that might profit from the developed optimised study design, which is characterised by a reduction of total samples from 120 to 14 per individual without any loss of information. Afterwards, they might be used to guide linezolid dose selection and might therefore help to improve individual therapy and outcome of serious infections in the critically ill.