Twenty years ago, as part of a review of the clinical pharmacology of mild analgesics (Beaver, 1965, 1966), I evaluated reports of those analgesic trials of dextropropoxyphene that appeared to satisfy the minimum methodologic requirements for a controlled clinical trial of analgesic efficacy. On reviewing reports of studies that have been published since then, I find little need to modify my evaluation of the efficacy of dextropropoxyphene that appeared in 1966, at least in respect to the effect of single oral doses: In summary, dextropropoxyphene (hydrochloride) is a mild oral analgesic which has proven superior to placebo in doses of 65 mg or more but which is of questionable efficacy in doses lower than 65 mg. The drug is definitely less potent than codeine: the best available estimates of the relative potency of the two drugs indicating that dextropropoxyphene is approximately 1/2-2/3 as potent. Likewise, dextropropoxyphene in 32-65 mg doses is certainly no more, and possibly less, effective than the usually used doses of aspirin or A.P.C. (aspirin/phenacetin/caffeine). In the interim, the efficacy of dextropropoxyphene has been the subject of a number of other critical reviews (Miller et al., 1970; Miller, 1977), commentaries (Kiplinger & Nickander, 1971; Lasagna, 1976), and even congressional hearings (Beaver, 1979; Moertel, 1979). More importantly, new controlled clinical trials involving dextropropoxyphene hydrochloride or napsylate have been reported, and some of these use more sophisticated design and analysis than those available in 1966. I will therefore discuss the results of those newer studies of apparently suitable scientific design that meet at least the minimum criteria for a valid clinical assay of analgesic activity (Beaver, 1965; Houde et al., 1965,1966; Wallenstein & Houde, 1975; Beaver, 1983), and I will comment on only a few of the studies included in my previous review (Beaver, 1966).