Prescribed daily doses (PDDs) of antiepileptic drugs (AED) (N03A ATC group) were recorded for drugs used in monotherapy or in combination therapy in the University Hospitals in Ostrava, Czech Republic and Huddinge, Sweden. Plasma concentrations were used as an indicator of the quality of treatment. PDDs were compared with the defined daily doses (DDDs) suggested by WHO in the ATC/DDD index 2005. Request and reply forms for therapeutic drug monitoring (TDM) were used as a source of mean PDDs. The study included 2,824 adult out- and in-patients in Huddinge treated from 1995 to 1999 and 1,268 out-patients treated in Ostrava from 1993 to 2004. The differences in PDD were tested by Student's t-test. Mean values of PDD were used when patients were examined more than once. Doses given in mono- and polytherapy were compared. Mean PDDs (in mg) in mono-/polytherapy in Huddinge and Ostrava were as follows (DDDs in parenthesis): carbamazepine 588/842 and 618/770 (1,000), clonazepam 3.0/2.5 and 3.4/2.4 (8), phenytoin 278/314 and 291/288 (300), gabapentin -/1,533 and -/921 (1,800), lamotrigine 228/228 and 216/195 (300), phenobarbital 90/75 and 183/117 (100), vigabatrin -/1,794 and -/1,259 (2,000), valproic acid 1,139/1,476 and 814/950 (1,500). The PDDs of most of the AEDs were lower than the DDDs with the exceptions for valproic acid (Huddinge, in polytherapy only), phenytoin, for which PDDs and DDDs were very close, and phenobarbital for which they were similar in Huddinge but higher in Ostrava. PDDs in monotherapy were only slightly lower than in combination therapy. Patients with plasma concentrations within the therapeutic range were usually treated with slightly higher doses than the remainder. In general, plasma concentrations tended to be in the low therapeutic range. The differences in PDDs between hospitals were significant in the case of valproic acid (P < 0.001), phenobarbital (except monotherapy within), vigabatrin, and gabapentin (P < 0.01), and carbamazepine (in monotherapy P < 0.05, polytherapy P < 0.01). Our data suggest that the DDDs of AEDs should be reconsidered as, in the majority of cases, they appear to be too high.