We have recently described a novel assay to assess ex vivo the activity and selectivity on cyclooxygenase-1 and -2 (EC 1.14.99.1) of non-steroid anti-inflammatory drugs (NSAID) administered to rats [Br. J. Pharmacol. 126 (1999) 1824.]. Here, we have extended these studies to humans. Healthy male volunteers were given orally one of the following drugs (mg) for 5 days: etodolac (200 or 400 b.i.d.), meloxicam (7.5 or 15 q.d.), nimesulide (100 or 200 b.i.d.), nabumetone (500 or 1000 b.i.d.) or naproxen (500 b.i.d.). Blood samples were withdrawn from the volunteers before and up to 24 h after the last dose. Plasma obtained from the blood was tested for its ability to inhibit prostanoid formation in interleukin-1 beta -treated A549 cells (cyclooxygenase-2 system) and human washed platelets (cyclooxygenase-1 system). Plasma from etodolac-treated subjects demonstrated a slight selectivity towards the inhibition of cyclooxygenase-2. This effect was more prominent in plasma from subjects receiving meloxicam or nimesulide. Plasma from nabumetone-treated subjects showed no or little selectivity towards cyclooxygenase-1 depending on the dose of drug administered, while plasma taken from subjects receiving naproxen was more active at inhibiting cyclooxygenase-1 than cyclooxygenase-2. In conclusion, we have demonstrated that this assay can be used to assess ex vivo the relative activity against cyclooxygenase-1 and cyclooxygenase-2 of NSAIDs consumed by human volunteers. It is to be hoped that data from such systems will aid in our understanding of the relationships between the differential inhibition of cyclooxygenase-1 and cyclooxygenase-2 by NSAIDs and their reported efficacies and (gastrointestinal) toxicities. (C) 2001 Elsevier Science B.V. All rights reserved.4264167195103