The purpose of this study was to evaluate a protocol which enables determining luminal drug concentrations after oral drug administration in man. Human intestinal fluids were aspirated from two sampling sites (duodenum and jejunum) at different time points after oral intake of theophylline; an immediate- and a slow-release dosage form were used to demonstrate the feasibility of discriminating between different formulations. Osmolarity and pH of the aspirates were measured and theophylline concentrations were determined by HPLC-UV. After intake of the immediaterelease formulation of theophylline, duodenal maximum concentrations up to 3 mM were reached within 30 min. Theophylline appeared to be almost completely absorbed before it reached the second sampling site in the jejunum, as observed jejunal concentrations were lower than 10% of the maximal duodenal concentrations. These results are in agreement with fast dissolution and fast absorption through the intestinal mucosa, which could be expected as theophylline belongs to class I of the Biopharmaceutical Classification System. In contrast to the immediate-release formulation, administering the slow-release dosage form resulted in a gradual appearance of theophylline, reaching maximal intestinal concentrations below 300 muM. The proposed methodology can be used to assess luminal drug concentrations and to monitor the time- and site-dependent composition of intestinal fluids after intake of an oral dosage form. This approach may contribute to a better understanding of the behaviour of oral drug formulations in the gastrointestinal tract and may be exploited to further unravel the complexity of the gastrointestinal absorption process. In addition, knowledge of luminal drug concentrations may assist in the selection of drug concentrations applied in in-vitro permeability assays.