The bias in relative risk estimates resulting from misclassification of outcome status has been shown to be a function of sensitivity and specificity of the classification procedure and of disease frequency. In epidemiologic studies, it may not be possible to obtain estimates of sensitivity and specificity. This situation frequently arises in studies of risk factors for coronary artery disease, where invasive procedures may be necessary to validate diagnosis. However, an estimate of the predictive value of a positive result in the classification procedure may be far more readily obtained than estimates of sensitivity and specificity. In this paper, exact and approximate formulae for the adjusted relative risk in terms of the predictive value of a positive test are derived. Hypothetical examples of epidemiologic studies of coronary artery disease are provided to illustrate the use of these formulae.