This study examines the relationship between serum and saliva topiramate concentrations, and attempts to determine if saliva may be a useful alternative to serum for therapeutic monitoring. Saliva and blood specimens were collected from 31 epilepsy patients (mean age 10.5 +/- 6.0 years; range 2.5 years to 24.8 years), and topiramate concentrations were determined by fluorescence polarization immunoassay. One patient's results were omitted because the saliva concentration was below the limit of quantitation of the assay. A strong correlation exists between serum and saliva topiramate concentrations (adjusted r(2) = 0.97, n = 30, P < 0.0001). The mean fraction of saliva to serum concentration is 89.8% +/- 12.1% (range 62.9% to 112.7%). The results of this study support the use of saliva as a viable alternative to serum for monitoring topiramate therapy. Topiramate concentration in saliva: an alternative to serum monitoring.