Arteriovenous differences for lactate and glycerol reported across the human forearm are inconsistent in direction and magnitude. Such inconsistency could represent the effects of differing forearm compositions. The hypothesis was tested by examination of 37 studies of forearm arteriovenous differences for lactate, glycerol, glucose, and oxygen (only measured in 25 studies) in 23 normal subjects after overnight fast. In 16 studies, glycerol was taken up rather than released by the forearm, and in 12 of these the arteriovenous difference was greater than could be accounted for by analytical variation. The arteriovenous differences for glycerol and lactate were positively correlated (r = .44, P less than .01). The hypothesis that glycerol and lactate uptake might reflect a more "oxidative" forearm was not borne out, since neither glycerol nor lactate arteriovenous differences correlated with that for oxygen, although oxygen and glucose arteriovenous differences and fluxes were correlated (fluxes: r = .60, P less than .01). The arteriovenous difference for glycerol was positively related to body mass index, arguing against a variable contribution from fat. The hypothesis that the direction of glycerol and lactate exchange would reflect the forearm composition was not borne out by the analysis of repeated studies on the same individual, which showed that the variation within subjects was not significantly less than that between subjects. Therefore, we conclude that in approximately 40% of studies in normal subjects after an overnight fast, the forearm will show glycerol uptake, although we have been unable to identify any physiological reason for this phenomenon. Peripheral glycerol uptake has implications for studies in which glycerol release is taken as a measure of lipolysis.