This paper reports the effects of a tailored Web-based delivery system on self-efficacy as it relates to a patients' response to acute myocardial information (AMI) symptoms. The data reported are from MI-HEART, a randomized trial examining ways in which a clinical information system can favorably influence the appropriateness and rapidity of decision-making in patients suffering from symptoms of acute myocardial infarction. Participants were randomized into one of three groups: tailored Web-based, non-tailored Web-based and non-tailored paper based. A theoretically based behavioral-cognitive model was used to identify key variables upon which to tailor education material. A key variable in the model is self-efficacy, operationalized with a three-dimensional scaling. Results show trends in improved self-efficacy scores for all groups at 1-month follow up, with sustained significant increases in baseline to 3-month scores only in the tailored Web-based group. One possible explanation could be related to "hit-count", which was significantly higher in the tailored group. This study is a first step in quantifying the contribution of Web-based tailoring over non-tailoring in changing key determinants of patient delay to AMI symptoms.