Levetiracetam (LEV) is a new antiepileptic drug approved as add-on therapy. Previous studies indicated that LEV has no relevant interactions with other antiepileptic drugs. The aim of this study was to investigate the influence of LEV dose, age, and co-medication on the serum concentration of LEV. In total, 363 samples of 297 inpatients who fulfilled the inclusion criteria (e.g., trough concentration, body weight available) were investigated. A patient was considered twice only if his co-medication had been changed. The LEV serum concentration in relation to LEV dose/body weight [level-to-dose ratio, LDR, (microgram/mL)/(mg/kg)] was calculated and compared for the most frequent drug combinations. Analysis of covariance (using age as covariate) carried out on the log-transformed data showed that co-medication had a highly significant (P < 0.001) effect on LEV serum concentrations. The median LDR of LEV was 0.32 for LEV + phenytoin, 0.32 for LEV + carbamazepine, 0.34 LEV + oxcarbazepine, 0.45 for LEV + lamotrigine, 0.46 for LEV + phenobarital, 0.52 for LEV monotherapy, 0.53 for LEV + valproic acid, and 0.54 LEV + valproic acid + lamotrigine. In co-medication with phenytoin (P < 0.001), carbamazepine (P < 0.001), and oxcarbazepine (P < 0.004), the LDR of LEV was significantly lower than it was with LEV monotherapy, whereas the LDR of LEV of patients on co-medication with valproic acid or lamotrigine did not differ significantly from the LDR of LEV of patients on LEV monotherapy (P > 0.05). Regression analysis including all 363 samples confirmed that other drugs (e.g., phenytoin, carbamazepine) lower LEV concentrations. In addition to co-medication, age had a significant effect on clearance of LEV. Children had lower LEV concentrations than adults on the same LEV dose per body weight. In contrast to other studies, our data point out that other enzyme-inducing antiepileptic drugs (e.g., phenytoin, carbamazepine) can moderately decrease LEV serum concentrations (by 20-30%). However, our observations should be confirmed by prospective pharmacokinetic studies.