Since its introduction in the late 19th century, the so-called cohesion theory has become widely accepted as explaining the mechanism of the ascent of sap. According to the cohesion theory, the minimum standing vertical xylem tension gradient should be 0·01 MPa m−1. When transpiration is occurring, frictional resistances are expected to make this gradient considerably steeper. The results of numerous pressure chamber measurements reported in the literature are generally regarded as corroborating the cohesion theory. Nevertheless, several reports of pressure chamber measurements in tall trees appear to be incompatible with predictions of the cohesion theory. Furthermore, the pressure chamber is an indirect method for inferring xylem pressure, which, until recently, has not been validated by comparison against a direct method. The xylem pressure probe provides a means of testing the validity of the pressure chamber and other indirect techniques for estimating xylem pressure. We discuss here the results of concurrent measurements made with the pressure chamber and the xylem pressure probe, particularly recent measurements made at the top of a tall tropical tree during the rainy season. These measurements indicate that the pressure chamber often substantially overestimates the tension previously existing in the xylem, especially in the partially dehydrated tissue of droughted plants. We also discuss other evidence obtained from classical and recent approaches for studying water transport. We conclude that the available evidence derived from a wide range of independent approaches warrants a critical reappraisal of tension-driven water transport as the exclusive mechanism of long-distance water transport in plants.