In Western Europe medicine packages contain an insert prepared by the manufacturer which enumerates the drug side-effects. We investigated the influence of this insert on alleged theophylline side-effects. Forty literate adult asthmatics were randomly allocated into two groups (n = 20 each): theophylline packages contained the manufacturer’s insert in group A but not in group B. Theophylline was prescribed (10 mg.kg-1 body weight qd) for one week. During this period the patients filled a diary grading 13 different symptoms from 0 to 3; 5 of these symptoms were listed on the insert as theophylline side-effects. On the eighth day the patients were interviewed and theophylline blood levels measured. Theophylline side-effects were significantly more marked in group A than in group B, whereas the other symptoms were of similar magnitude. Eight patients prematurely stopped their treatment in group A vs 3 in group B, because of alleged intolerance. Theophylline blood levels did not differ significantly in the two groups; neither did they in the subgroup which stopped treatment and in the one which complied to prescription. We conclude that side-effects were suggested to the patients by the insert and/or that the insert increased their awareness of side-effects, with a subsequent detrimental influence upon compliance to therapy.