MIKE Davies is concerned by the widespread off-label use of medicines by veterinary surgeons, and cites the use of oral tramadol as an analgesic for osteoarthritis as an example ( VR , June 30, 2012, vol 170, p 680). He is right to express concern about the possible side effects and lack of efficacy of unlicensed medicines, but – like the medicines regulators themselves – appears to see less than half the picture. The current licensing system has major drawbacks. Three that are relevant are: ▪Only those medicines likely to make a substantial profit for a manufacturer will become licensed, thus introducing a degree of arbitrariness to which drugs will become licensed. There are many therapeutic niches where useful medicines exist that would improve the welfare of animals, but which are unlikely to make a profit for a manufacturer and so are unlikely to be licensed under the current system. Perhaps one such niche is a ‘secondary’ oral analgesic – that is, in addition to a licensed NSAID, or instead of an NSAID in those cases where NSAIDs are contraindicated – for long-term home use for conditions such as osteoarthritis. This is where unlicensed opioids such as tramadol are currently used. ▪The licensing system guarantees that the licensed analgesics have at least some efficacy and tells us the more common side effects they have (it misses some of the rare, and, indeed, some common, side effects). That is valuable and …