The relationship between levels of written drug therapy information provided to patients and the degree to which patients comply with their antibiotic regimens was studied. The study included 80 adult patients who received a prescription for a ten-day course of ampicillin, phenoxymethyl penicillin potassium or tetracycline and who had their prescriptions dispensed by a specific pharmacy. Two levels of information were used: (1) the usual prescription label and (2) the usual prescription label, plus an auxiliary label instructing the patient to consume the medication for the full course of therapy, plus an information sheet explaining the rationale for consuming the medication at the proper times and for the full course of therapy. The level of information was randomly assigned for the first week and alternated each week thereafter for a total of four weeks. Patients were randomly assigned to unannounced home interviews on the third, sixth or ninth day of therapy. The measure of compliance was a physical count of the remaining medication as compared to the count which should have been present if all medication has been consumed correctly. The factors included in the experimental design were level of information, week of the experiment (nested in information level) and duration of therapy. The only factor significantly (α = .05) associated with compliance was information level. The mean percent compliance rate for those who received the higher level of information was 85%, while the rate for those who received only the usual prescription label as information was 63%. One implication of these findings is that the pharmacist can make a valuable contribution to the health care of his patients through effective counseling with written information.