For patients taking two or more medications concurrently, interactions among the drugs can cause undesirable effects or negate desired responses. In modern pharmacy practice, an important role of the pharmacist is to detect potentially harmful interactions and take appropriate action to prevent their occurrence. Pharmacy computer systems offer potential for improving pharmacists' effectiveness in the detection and followup of drug interactions. Based on a survey of southern Michigan pharmacists, relationships between computer use and pharmacists' attitudes and activities in drug interaction monitoring were investigated. Respondents included users of two major computer systems as well as pharmacists who do not use computers. Results suggest that general statements cannot be made about the effect of computer use on drug interaction detection. Users of one of the two computer systems detected and followed up on interactions more frequently and were more likely to report improved knowledge of drug interactions than non-users. Frequencies of drug interaction detection and other related measures reported by users of the second computer system were similar to those for pharmacists not using computers. Computer system characteristics which might lead to these differences are discussed.