Adenosine triphosphate (ATP) has been suggested as a mediator of active hyperemia and its levels have been reported to increase in the venous plasma from contracting skeletal muscle. However, the source of the ATP is unknown. The present study indicates that a large portion of the plasma ATP is released from the formed elements of blood when the blood is collected in the presence of EDTA. When EDTA was added to blood that was previously incubated at 37 degrees C for 5 min to destroy all free ATP, the ATP level was 0.57 plus or minus 0.12 (plus or minus S.E.) nmoles/ml. However, it was possible to detect exogenously added ATP only when blood samples were collected into EDTA; collection into saline or citrate afforded no protection against ATP degradation by the ATPases of the blood. In dog hindlimb preparations perfused at constant flow or constant pressure, the venous plasma ATP of blood collected in the presence of EDTA exhibited no consistent increase during or following tetanic contraction of the muscles. In isolated, perfused rat hindlimbs, no ATP was detectable in the venous effluents from resting or contracting muscles (ATP smaller than 0.08 nmoles/ml). However, the levels of adenosine in the venous effluents were greater in contracting than in resting hindlimbs. The data indicate that it is not possible to make valid determinations of plasma ATP levels and thus, one cannot determine the role of ATP in active hyperemia based on these data. However, the currently available data from isolated muscle preparations do not support the concept that ATP is released from contracting skeletal muscle, and therefore, it is unlikely that ATP is a mediator of the metabolically-linked local regulation of skeletal muscle blood flow. The enhanced release of adenosine from contracting rat hindlimb muscles may indicate a role for this nucleoside in the regulation of blood flow in skeletal muscle.