Advancing our understanding of pain mechanisms and the need for improved analgesic treatments faces challenges in both clinical and laboratory domains. The preclinical approaches provide advances in our fundamental understanding of the neurobiology of pain, but the gaps between molecules and pathways to the patients need to be addressed. Viewpoints differ on the notion that animal models are the culprits of the failure to produce new pain drugs. 1 But perhaps, it is not a fault of the models, but the interpretation of the information provided by them? Pain measured in humans often relies on an analogue scale (ie a standard rating scale of 0–10). By definition, the threshold would lie around 2 and this is what is measured by reflex responses in animals. Consequently, a drug that is effective on threshold measures is likely to fail when confronted by the pain levels of 6–7 that patients in trials report. Here, a different approach to preclinical investigation of drug efficacy is needed. We confess to being besotted by neuronal responses obtained by in vivo electrophysiology, but these do provide an unbiased, objective measure of low and suprathreshold multi-modal responses in pain models. Indeed, these electrophysiological measures of neuronal activity likely equate better to clinical pains. 2 But then again, we might be geeks....

We would argue that done well, preclinical studies can be strong predictors of drug efficacy in the clinic; identification of cyclooxygenase-2 (COX-2) blockers, triptans, anti-nerve growth factor (NGF), and anti–tumour necrosis factor a (TNFa) therapies are all examples of successful translation