Young rhizome sprouts of the herbaceous perennial Jaumea carnosa were propagated from material collected in a salt marsh along the central California coast. The sprouts were transplanted to flats of sand sown with different densities of seeds of a representative glycophyte, Lolium perenne L. "Derby, turf type." Controls flats contained only Jaumea or Lolium. Three series of replicated flats were watered from above with dilutions of seawater in 1/10 strength Hoagland solution, such that dissolved salts were 400, 4000 or 11,600 ppm. Two other series were continuously subirrigated with 400 or 11,600 ppm salt water. After 61 days of treatment in a greenhouse with a 30/11°C thermoperiod (mean daily max/min), all plants were harvested and weighed. In the monospecific control flats, the growth of both species declined with increasing salinity, but the relative decline of Lolium was three times that of Jaumea. Jaumea's root: shoot ratio was also less affected by salinity. Both species grew well when subirrigated by 400 ppm salt water, but grew poorly when subirrigated by 11,600 ppm salt water, indicating that aeration alone is not the most significant factor in the marsh. The effect of interspecific competition on Jaumea was marked at low salinity, depressing growth by 52% compared to controls, but at high salinity the competitive effect was insignificant, whether the plants were watered from above or subirrigated. This supports the hypothesis that intolerant halophytes such as Jaumea are restricted in nature to salt marshes because they are poor competitors with glycophytes on non-saline soils.