Abstract Background Accurate quantification of xylem sap ABA concentrations is important to underpin models of root-to-shoot ABA signalling to predict the physiological effects of soil drying. Growing tomato plants in a whole plant pressure chamber allowed sequential xylem sap collection from a detached leaf, the petiole stub of an otherwise intact plant and finally the de-topped root system of the same plant, to determine the impact of xylem sap sampling methodology on xylem ABA concentration. Since xylem sap can contain bound forms of ABA, a novel gas chromatography-mass spectrometry (GC-MS) procedure was developed to chemically separate free ABA from two in planta bound ABA forms known as Adducts I and II and ABA-glucose-ester (ABA-GE). Results Xylem sap ABA concentrations were highly dependent on the sampling methodology used: the highest concentrations were detected in sap collected by applying an overpressure to detached leaves following the measurement of leaf water potential. Irrespective of xylem sap source, the wild-type cultivars Ailsa Craig and Rheinlands Ruhm had higher free ABA concentrations than a range of ABA-deficient mutants (notabilis, flacca and sitiens). However, in the mutants, concentrations of bound forms of ABA were similar to wild-type plants, and similar to free ABA concentrations. Conclusions Although xylem concentrations of these bound ABA forms and ABA-GE suggest they have a limited physiological impact on ABA homeostasis in tomato, the methods developed here will allow a more complete understanding of ABA biochemistry and root-to-shoot signalling in species known to have higher concentrations of these compounds.