Admission to clinical trials is often based on the assumption of homogeneity of the population. A group of 60 patients, all with pain in the neck or shoulder of at least 3 months duration, were studied. Expectation was graded before treatments were started by sharing out 100 points between freedom from side effects, pain relief, depression relief, improved mobility, improved sleep and speed of action. A double-blind crossover study of two established anti-inflammatory analgesics and placebo was carried out. Using analogue scales, patients were asked to grade their response. Side effects were recorded, and preference was established at the end of the study.Although all the patients were in sufficient pain to require medical attention, some rated relief of depression, improvement in sleep or lack of side effects as more important than pain relief. Differences between drug and placebo were most clearly demonstrated in those patients whose main concern was improved mobility. In our view it is important to select patients who are in need of a dominant property of a drug for a trial of this property and this may have ramifications across the medical spectrum.