Greater impairment at higher levels was demonstrated for five of 17 measures. Aman et al. conducted a series of studies in children using the high versus low dose design (39-41). The effects of VPA on psychomotor performance were studied in 46 children who were on monotherapy and had been seizure-free for at least 2 months (39). In addition to baseline testing, the children underwent testing before the morning VPA dose and 1 week later again -2.5 h after the morning dose. The analysis also included a comparison between patients receiving a VPA dose above or below 20 mgkg per day. Patients who were on a lower dose performed at a higher level on 4 of 26 measures, but time of testing in relation to the VPA dose had no significant effect on performance. The latter observation may be related to the study design. VPA was administered as syrup in eight patients and as tablets in 38 patients. No levels are provided in the report, and it is not specified whether tablets were en- tericcoated. After intake of entericcoated VPA tablets, VPA absorption into the blood may not begin until several hours later, in which case the VPA level continues to decrease after intake until absorption exceeds elimination. Thus, in many patients in this study, levels 2.5 h after the morning dose may have been actually lower than before the morning dose. In a study based on a very similar design, the neuro psychological effect of CBZ monotherapy was studied ifi 50 children with well-controlled seizures (40). The children were tested once shortly after they had taken their CBZ dose, and once before their dose. The correspond- ing mean CBZ saliva concentrations were 10.3 and 5.3 yM, respectively. The patients performed significantly better during the postdose session than during the pre- dose session on four measures (seat activity, attention span, motor steadiness, and task specific response times), with an opposite difference for two measures of response speed. No levelhesponse correlations with actual CBZ concentrations across subject could be demon strated in this study. The same group of authors used an almost identical protocol to test the effect of PHT on cognitive motor performance in children (41). PHT concentrations and time of testing in relation to PHT dose had few discernible effects on psychomotor performance, despite documented PHT level fluctuations of the order of 50%.