The Borg Ratings of Perceived Exertion (RPE) Scale has proven to be a highly popular instrument in measuring the subjective responses of individuals to a given work or exercise task. Historically, the instrument was designed to correlate highly with the heart rates in young-to-middle-aged men performing various tasks. The body of literature, however, has revealed inconsistencies in the extent of just how strong the relationship is between ratings of perceived exertion and various physiological criterion variables, most notably, heart rate. In addition, most studies have invoked the question of whether the criterion-related validity coefficients derived from the relationship between ratings of perceived exertion and a specified physiological criterion variable are just as valid as those for which the Borg RPE Scale was originally performed. A meta-analysis, therefore, was undertaken to determine the magnitude of the relationship between ratings of perceived exertion scores and each of three commonly used physiological measures or criterion variables: heart rate, blood lactate, and oxygen uptake. Results show that by using Tests of Homogeneity for each physiological criterion variable, the observed sample size-weighted validity coefficients are heterogeneous. The median of the mean sample size-weighted validity coefficients is .574 for heart rate, .561 for blood lactate, and .480 for oxygen uptake. Each study in the meta-analysis was grouped by the study characteristics of subject gender, fitness level, RPE Scale, exercise type, exercise protocol, and study quality. For heart rate, the highest validity coefficients are those in which the subjects are highly fit, the exercise type is fairly unusual, such as swimming, and the subjects are required to maximally exert themselves. For blood lactate, the highest validity coefficients are for females, healthy-inactive subjects, the 15-point RPE Scale, treadmill use, and swimming. For oxygen uptake, the highest validity coefficients between ratings of perceived exertion and oxygen uptake are for swimming. In a meta-analysis of study effects, when the validity coefficients are analyzed by study, the resultant mean validity coefficients are only somewhat higher (ratings of perceived exertion and heart rate, .657; ratings of perceived exertion and blood lactate, .642; ratings of perceived exertion and oxygen uptake, .609) than those obtained using sample size-weighted validity coefficients. Finally, corrections for bias generally resulted in increased validity coefficients and decreased variances.