AbstractThe goal of this study was to identify the most important variables affecting bioluminescent ATP, ADP and AMP measurements in plasma and to develop an assay that takes these variables into account. Blood samples were drawn from conscious dogs. A ‘stop solution’ containing EDTA was prepared, which greatly retarded plasma ATP degradation by chelating Mg+2 and Ca+2 that are co‐factors for many ATPases. Stop solution and blood were mixed using a two‐syringe withdrawal system. Samples were centrifuged twice in order to remove red blood cells, and ATP was measured in the supernatant using the firefly luciferase assay. Sample pH was adjusted to the optimal range (7.75–7.95) and Mg2+ (necessary for the luciferase reaction) was added back to the sample within the luminometer 2 s prior to luciferase addition. Four assay tubes were prepared for each plasma sample, containing standard additions of 0–15 pmol added ATP, in order to quantify native plasma ATP content. In separate plasma/stop solution samples ADP + ATP was measured after converting ADP to ATP via the pyruvate kinase reaction, and AMP + ADP + ATP was measured after addition of both myokinase and pyruvate kinase. Addition of forskolin and isobutylmethylxanthine (IBMX) to the stop solution to inhibit platelets resulted in lower ATP concentrations. Measurement of ATP and haemoglobin from lysed erythrocytes revealed that haemolysis exerts a strong influence on plasma ATP concentration that must be taken into account. Copyright © 2003 John Wiley & Sons, Ltd.