Woolf et al have recently called for the development of a mechanism-based pain taxonomy to guide the individualization of treatment based on each patient's pain mechanisms. Although any scientific physician could endorse this ideal, small academic clinical trials so far have failed to identify obvious differences in the response of different pain symptoms in the same condition to various drugs. In contrast, there are clear differences in the analgesic responses of patient groups distinguished on the basis of etiology or tissue origin of pain, factors which tend to be associated with groups of mechanisms. The few tests to diagnose pain mechanisms remain too delicate, time-consuming, or uncomfortable for general clinical use. To understand how best to exploit new mechanistic insights to assign treatments, one must scrutinize the relative value of diagnostic classifications based on etiology, tissue, and individual patients' pain characteristics in large clinical trials. Research priorities should include developing simple methods for assessing pain mechanisms in the clinic and increasing the efficiency of pain assessment methods in clinical trials. I describe a collaborative research agenda for academic pain researchers and funding agencies, the pharmaceutical industry, and regulatory bodies.