Within the past 20 years many cyclooxygenase (COX) inhibitors were removed for unwanted drug effects shortly after entering the drug market (e.g. benoxaprofen and isoxicam), whereas others (e.g. diclofenac and ibuprofen) were not. This has continued with the suspension of the sale of the COX-2 inhibitors rofecoxib, valdecoxib and lumiracoxib, whereas others (e.g. celecoxib and etoricoxib) are still available. All these compounds share the same molecular mode of action but differ considerably in their pharmacokinetics. Determination of pharmacokinetic-pharmacodynamic relationships should help to pinpoint deficits, answer pending questions and lead to a safer drug use. Here, we provide evidence that applying the ex vivo human whole-blood assay could provide a valuable tool for defining the lowest effective dose and the adequate dosing interval of COX inhibitors. In our opinion, such an approach could reduce unwanted drug effects and obviate drug removals.