We thank Ruan et al. for recognizing our efforts to establish an effective meloxicam dose for inflammatory-based pain prevention in mice. Our study found that although 5 or 20 mg/kg almost completely prevented inflammation, neither dose resulted in effective pain control. Ruan et al. make the point that inflammation is needed to activate the peripheral opioid system. They argue that if some inflammation had been allowed to persist, it may have resulted in greater intrinsic pain control, in which case meloxicam might have been shown to have efficacy at the relatively lower dose rates (0.5-2.5 mg/kg) usually prescribed to humans. Firstly, such comparisons are meaningless until there is a sufficiently sophisticated method of allowing animals to report on their pain experience or the affective properties of drugs. But more importantly, what Ruan et al. appear to have overlooked is that our dose range included 1 mg/kg; which allowed inflammation to develop to a point that was less severe than in untreated mice but more severe than in those receiving 5mg/kg meloxicam (see Figs. 1 and 2 of the original manuscript). There was also a generally uniform relationship between the meloxicam dose rate and inflammation that contrary to their hypothesis was not reflected in the pain metrics. It is important to remember that a main aim of our study was to investigate the apparent lack of efficacy of meloxicam in mice relative to rats. We cite several studies where mice of various strains failed to respond to meloxicam in the suggested dose range (those by Wright-Williams et al., Roughan et al. Miller et al. and Matsumiya et al.). Since an obvious explanation is that these used sub-optimal pain detection methods, in addition to imaging and using automated behaviour recording and analysing bodyweights, in the current study, we hoped to benefit from using the Mouse Grimace Scale (MGS) and its growing reputation as currently one of the most effective methods for assessing pain and analgesic efficacy in both mice and rats. The dose range tested came from an extensive series of investigations on the effects of between 2 and 20 mg/kg meloxicam for post-laparotomy or post-vasectomy pain prevention in mice (Wright-Williams, 2008). Here, we wished to explore whether the largely negative findings could have been because meloxicam did not effectively prevent inflammatory-based pain. Since publishing the current findings, we have repeated the inflammation imaging and MGS analysis, but included the same manual analysis of pain-specific behaviour we originally used to show meloxicam is effective in rats. However, the study found that 10 mg/kg meloxicam, 20 mg/kg tramadol or giving both was still not effective in BALB/c mice undergoing laparotomy. Similar behaviour analysis indicates meloxicam within the dose range Ruan et al. suggest is also not effective for ovariohysterectomy in rabbits (Leach et al., 2009), and the only studies we are aware of that have described positive results from using meloxicam at between 2 and 3 mg/kg in mice used ot...