This work was carried out to adapt the electrolyte leakage technique todurum wheat and then to evaluate its relevance in the assessment of the cellmembrane stability as a mechanism of water stress tolerance in this species.Themethod currently used is based on in vitro desiccation ofleaf tissues by a solution of polyethylene glycol (PEG) and a subsequentmeasurement of electrolyte leakage into deionised water. It consists of threesuccessive steps: (1) a washing treatment to remove solutes from both leafsurfaces and cells damaged by cutting; (2) a stress period during which theleaftissues are plunged in a PEG-solution and (3) a rehydration period during whichafter-effects of the stress are evaluated. During the washing period, the majorpart of electrolytes was removed within 15 min. Varying the stressconditions influenced both the percent and the kinetics of electrolyte leakageduring rehydration. Electrolyte leakage exhibited a characteristic patternreflecting the condition of cellular membranes (repair and hardening). Inpractice, we recommend a 15-minute washing time, a10-hour stress period and 4 h of rehydration. Theextent of the cell membrane damage not only correlated well with the growthresponses of wheat seedlings belonging to various cultivars to withholdingwaterbut also with the recognised field performances of these cultivars. Therelativeproportion of endogenous ions lost in the effusate during the rehydration stepmay vary strongly according to the element analysed and the precise nutritionalstatus of the plant should therefore be considered. However, an increase ininorganic ion leakage does not fully explain the recorded PEG-induced increasein electrical conductivity (EC) during the subsequent rehydration step andorganic ions are probably also involved in such an increase.