Phentolamine (30 mg) was administered intravenously to subjects with idiopathic chronic low back pain in a novel placebo-controlled test to determine whether this α-adrenergic antagonist would reduce their pain. The effects of infusions on spontaneous pain and stimulus-evoked pains (touch, cold, tapping and deep pressure) were evaluated separately. All subjects gave strong placebo responses (reduced pain) that prevented assessment of specific drug effects. The placebo responses had onset latencies of 15–60 min, developed slowly over the next 15–45 min and persisted for hours or several days. These results not only reinforce the understanding that placebo controls are essential in the evaluation of drugs or other palliative procedures on patients with chronic pain but also indicate that the control paradigms must allow for placebo effects that are slow to develop and very persistent.