Gastrointestinal (GI) damage by non-steroidal anti-inflammatory drugs (NSAIDs) is an important Public Health problem due to morbility and elevated mortality rate. The logic behind the development and sale of molecules which are selective cyclooxegnase-2 (COX-2) inhibitors called coxibs, is to limit the undesired effects of traditional NSAIDs, which should theoretically derive from the inhibition of COX-1. With respect to the emphasis of the initial trials, to now epidemiological studies, open-label studies, meta-analyses and reviews of the same data with longer follow-up, have produced opposite conclusions. Indeed, a recent series of meta-analyses has shown that the risks of GI events are similar for coxib and diclofenac, while they are significantly higher during the assumption of ibuprofen or naproxen. Moreover, the presumed lower gastrolesivity of coxibs is based on a highly simplified hypothesis, that the gastroprotective PGs derive from COX-1 and that the phlogistic processes are related to COX-2. Also in geriatric populations, though less tolerated than coxibs, diclofenac presents minor GI side effects when compared with naproxen and ibuprofen. In this context, in the case of moderate pain intensity, it is possible to use combinations with weak opioids, such as paracetamol-tramadol. Though intestinal damage by NSAIDs are a nosological entity of growing interest, to now no trial has been conducted with optimal criteria to demonstrate the superiority of coxibs over traditional NSAIDs. For this reason, chronic inflammation of the intestine still represents a contraindication to the administration of coxibs.