OBJECTIVES: To compare the hyperforin and hypericin content of currently available St. John's wort products and to determine their batch-to-batch reproducibility. DESIGN: Representative products were obtained either directly from the manufacturer or purchased from pharmacies in and around Frankfurt, Germany. For five batches from each of the eight manufacturers, 10 individual dosage forms (tablets or capsules) were analyzed for both hyperforin and hypericin content. SETTING: Laboratories of the Institute of Pharmaceutical Chemistry at Johann Wolfgang Goethe University, Frankfurt, Germany. PRODUCTS: Eight German St. John's wort products containing from 250 mg to 612 mg dry extract were studied. Three of these products are capsules, four are film-coated tablets, and one is a sugar-coated tablet. Two of the products (Jarsin 300 and Neuroplant 300) are also available in the United States. METHODS: Hyperforin concentrations were analyzed by high-performance liquid chromatography. Total hypericin concentrations were determined by polarography, an electrochemical method. Concentrations were compared among different batches of the same product and among products from different manufacturers. RESULTS: The products contained widely differing amounts of hypericin and hyperforin, even after correcting for differences in the amount of extract per dose. Some products demonstrated consistent concentrations of hyperforin and hypericin from batch to batch, others exhibited pronounced interbatch variability. CONCLUSION: The St. John's wort preparations studied exhibited large differences in hypericin and hyperforin content and are not interchangeable for the treatment of mild-to-moderate depression. Pharmacists should take this variability into account when counseling patients on the use of St. John's wort products.