Dextropropoxyphene is a weak opioid analgesic, widely used as a step 2 analgesic (according to WHO classification) in combination with peripheral analgesics, mainly paracetamol. Recent data have underlined its poor analgesic efficacy (in comparison with paracetamol), risks of serious adverse drug reactions (i.e. hepatic reactions, hallucinations, abuse, withdrawal symptoms, hypoglycaemia), possible lethality after overdose, its risk of accumulation in patients with renal failure or in elderly people and some pharmacokinetic insufficiencies (i.e. different half-lives for dextropropoxyphene and paracetamol). Taking into account these data, the drug committee of the Toulouse University Hospital (France) decided to withdraw dextropropoxyphene from the hospital formulary since 1 June 2005. The aim of our study was to investigate the consequences of this withdrawal by comparing use of analgesic drugs in Toulouse University Hospital before (2004) and after (2006) dextropropoxyphene withdrawal (using defined daily dose for 1000 hospitalization-days as the unit measure). Before withdrawal, dextropropoxyphene (in combination with paracetamol) was the second most used analgesic drug after paracetamol alone. After dextropropoxyphene withdrawal, total consumption of analgesic drugs decreased by 4.6% (2006 vs. 2004). There was a 28% decrease in consumption of step 2 analgesics [with an increase in oral tramadol and a slight decrease in codeine (in combination with paracetamol)]. During the same period, step 1 analgesic consumption increased by 11% (mainly paracetamol) and that of step 3 analgesics slightly decreased (-8%). These results show that dextropropoxyphene withdrawal was not associated with a marked switch in prescriptions towards other analgesic drugs. This paper underlines the interest of a hospital-based drug committee to promote rational drug use. Finally, the present data allow us to discuss putative misuse of dextropropoxyphene.