Objective: To explore the impact of the inclusion of a benefit message in a patient package insert on knowledge about medicines and on subjective benefit/risk perception. Setting: Female members of community social organizations, female relatives of psychology students, and caregivers to psychotic patients. Nature of the study: Randomized, controlled healthy human volunteer study with three parallel experiments, involving the inserts of cisapride, itraconazol, and risperidon. Design. Subjects were recruited in a convenience sample and randomized to one control and two intervention groups (one with a normal insert and one with an insert with a benefit message). Material and methods: Subjects were asked to read the inserts (using mock text in the control group) in 5 to 15 minutes. Knowledge of the medication was tested with 20 simple questions (to be answered Yes/No/Don't know) and benefit/risk perception with a five-point bipolar Likert scale. Results: In the three experiments respectively 89, 102, and 83 subjects were recruited. The provision of inserts increased the knowledge about medication in all the intervention groups. Thirty-one pet-cent, 41%, and 54% of the subjects who read a normal insert agreed that the benefit of the medicine was greater than its risks, compared to 62%, 64%, and 70% of subjects who read an insert with a benefit message included (P < 0.05 in all 3 experiments). Discussion: A hypothesis for further research is formulated: adding a section on benefit information within a patient package insert helps to integrate increased knowledge about medication into a more balanced benefit/risk perception.