Upon exposure to air at 20°C and 60% relative humidity (RH), an occlusion to water uptake developed in the basal stem segment of cut, flowering Sonia roses. The development of the occlusion was delayed by removing the leaves before storage, indicating an effect of transpiration. Placing the stem ends at 100% RH during exposure to air had no effect, which shows that water loss through the cut end is not the cause. Scanning electron microscopy did not show plant gums or tyloses in the water conducting elements.The hypothesis by Scholander et al. (1955) that water uptake into stems held in air and then placed in water occurs through conduit walls rather than the lumen was tested by using stems of which the cut surface was covered with laboratory grease and a ring of bark was removed (girdling). Girdled stems placed in water remained fully turgid when the girdled area was about 60 mm2or more. Water uptake was strongly inhibited when the girdled stems were exposed to air for 24–36 h, as in non‐girdled controls exposed to air for the same period. Inclusion of a surfactant (nonylphenoxypolyethoxy ethanol) in the vase water, however, facilitated water uptake after dry storage of normal cut flowering stems but did not improve water uptake into the girdled stem system, which is inconsistent with the hypothesis.It is concluded that the vascular occlusion developing upon exposure to air cannot be explained by decreased flow in the wall pathway for w