Efforts to restore semiarid wildlands in the western United States predominantly use fall seeding. Fall conditions are more amenable to seeding, and successfully over-wintered seeds or plants are poised to take full advantage of spring moisture. However, over-winter mortality can be a barrier to seeding success. One solution to avoid winter mortality without sacrificing the benefits of fall seeding is to delay germination of fall-sown seeds. At six northern Great Basin field sites over three planting years (18 trials), we tested whether a hydrophobic seed coating reliably delayed germination and increased seedling establishment of a native bunchgrass. Despite considerable variation among sites and years, seed treatment successfully reduced prewinter germination by 84% compared to untreated, but also consistently reduced final germination by 23%. Still, treatment resulted in an average of 2.1-fold higher emerged seedling density in seven trials, six of which were among the eight trials where more than 50% of untreated seeds germinated pre-winter. Emergence was greater for untreated seed in five trials, all of them in the same year, and all but one of them with below-average total germination. Our treatment consistently reduced pre-winter germination, but only improved emergence when pre-winter germination of untreated seeds was high. Continued research is merited with germination-delaying treatments and to experimentally define-and develop models that predict-winter-related mortality barriers. We also suggest future exploration of bet-hedging strategies that mix treated and untreated seeds where fall seeding is required and significant but variable risk of over-winter seedling mortality exists.