The Stress Gradient Hypothesis predicts facilitation will become more important than competition where abiotic stress is high, and the framework successfully predicts positive interactions between species in many systems. Fewer studies have focused on intraspecific facilitation, and to our knowledge none examine intraspecific interactions in Pacific Coast salt marshes of North America. We used two species that tend to occur in large, monospecific patches to test for intraspecific facilitation in a restored California marsh, where tides and evaporation during summer create moisture and salinity gradients across elevation. We tested performance of Frankenia salina or Jaumea carnosa in restoration plots using two treatments: clustered plantings to promote facilitation, and widely spaced plantings to limit interaction between individuals. We characterized the soil water potential gradient and measured plant survival, cover, physiology, and susceptibility to herbivory across elevation and planting treatments. Soil water potential declined sharply with elevation, suggesting plant stress should be higher upslope. However, tissue water potential was unaffected by elevation, and survival was high—suggesting growing conditions remained benign. A seawater addition treatment did not alter the response of plants or plant interactions to the abiotic gradient. At 19 months, intraspecific competition prevailed across the entire elevation gradient for both species, with less transplant cover in clustered plantings. However, clustered Frankenia fared better during a transient period of heavy rabbit herbivory. Aside from this temporary benefit, clustering did not reduce stress and strongly suppressed growth—suggesting that restoration designs for the high marsh should minimize competition.