This work reports the development and application of a negative ion mode online LC-ESI-MS method for studying the effect of water deficit on the carbohydrate content of Lupinus albus stems, using a porous graphitic carbon (PGC) stationary phase and an ion trap mass spectrometer. Using this method, separation and detection of several water soluble carbohydrates, ranging from mono-, di-, and oligosaccharides (raffinose, stachyose, and verbascose) to sugar alcohols was achieved in approximately 10 min. This on-line PGC-LC-ESI-MS method shows good linearity with correlation coefficients R(2)>0.99, selectivity, short analysis time, and limits of detection (LOD) ranging from 0.4 to 9 pmol for sugars and 4-20 pmol for sugar alcohols. This PGC-LC-ESI-MS method is sensitive and allowed us to detect even small alterations in carbohydrate levels in L. albus stems that resulted from a mild/early water deficit (nmol g(-1)DW). This paper describes details of our method and its application to the quantitative analysis of water soluble underivatised carbohydrates extracted from L. albus stem tissues that have been subjected to early and severe water deficit conditions, followed by a rewatering period.