To determine whether informed consent in a therapeutic trial modifies the analgesic effect of naproxen and placebo, we conducted a prospective, randomised, single dose, placebo-controlled trial. Patients were randomly selected to receive or not information concerning the study. All patients included were then given a single dose of naproxen and placebo according to a crossover, double-blind design. Forty-nine patients with mild or moderate cancer pain which did not need narcotic analgesics entered the study. Twenty-five received both treatments without any information and constituted the uninformed group. Twenty-four had a complete information about the trial; six refused to participate. The 18 others constituted the informed-consent group. Visual analogue scales of pain before and 30, 60, 120 and 180 min after the intake of naproxen and placebo were recorded. As an analgesic, naproxen was more effective than placebo in both groups of patients (p = 0.001). For naproxen as well as for placebo, the analgesic effect was better in the informed-consent group compared to the uninformed group (p = 0.012). The difference in therapeutic activity between naproxen and placebo was moderately higher in the uninformed patients (p = 0.08). We concluded that, in contrast with parallel studies, giving information in a crossover, placebo-controlled trial may increase the apparent efficacy of both the tested agent and the placebo, and decrease the perceived difference the two.