&NA;This pilot study investigated epidemiologically the potential for clinically significant drug‐drug interactions, a subclass of adverse drug responses, in two homes for the elderly. The agent (drugs). host (residents), and environment (rural and urban nursing homes) in the illness known as drug‐drug interaction were studied. Drug profiles on 188 subjects were compiled and screened by computer. Of these, 100 (80 in the rural home and 20 in the urban home) had the potential for clinically significant drug‐drug interactions. Laboratory diagnostic procedures essential to confirm the unintentional interaction of two or more drugs were beyond the scope of the study; therefore, physicians independently reviewed the records of 66 residents who could be interviewed, agreeing that 27 residents were potential drug‐drug interactors. These subjects had twice as many drug products and three times as many pro re nata products prescribed for them as did subjects without the potential for drug‐drug interaction. The possible cause of interaction found most frequently was digitalis combined with thiazide or furosemide. The results of this pilot study suggest several implications for practice within the research setting.