This study describes the development and testing of a high blood pressure protocol for use in emergency departments (ED) to enhance detection of those patients appropriate for subsequent referral. The protocol involves two serial blood pressure measurements and a patient interview to determine: 1) previous history of high blood pressure (HBP), 2) treatment in past year for HBP, and 3) usual source of medical care. The accuracy of patient reporting was validated by comparison with the patients' hospital record (reflecting outpatient and inpatient visits). Results indicate that these self-reports have high levels of sensitivity (range 90-100%) and specificity (range 79-96%). Use of the additional patient information increased the sensitivity of the screening protocol in identifying when and where a patient should be referred. Use of this methodology indicates that the protocol is a simple and effective method for HBP screening. The findings also suggest that the ED is an ideal site for screening the "hard-to-reach" hypertensive population.