Desert rodents in the western USA cache enormous numbers of seeds. Seed caches represent a reliable, although highly variable, source of seedling recruitment for numerous desert plants. Prefire caching of refractory seeds (i.e., those capable of forming persistent seed banks) is an important source of postfire seedling recruitment, but it is likely that prefire caches of short-lived seeds also may contribute to postfire regeneration. The primary objective of this study was to quantify seedling recruitment from artificial caches of 2 species with short-lived seeds planted after a stand-replacing wildfire in a singleleaf pinyon (Pinus monophylla) forest. I investigated the survival of 180 artificial caches of Yucca brevifolia and Prunus fasciculata. Caches were composed of 2, 4, or 8 seeds. Also, because soils of the burned area have unusually high coarse fragment volumes (59%), I examined, in greenhouse trials, how soil coarse fragment volumes (CFVs) impacted seedling establishment of these 2 species from caches. In greenhouse trials, the number of P. fasciculata seedlings increased with increasing CFVs. In contrast, Y. brevifolia seedling numbers decreased with increasing CFVs. In field transects, 17% of cached P. fasciculata seeds produced seedlings, whereas 12% of Y. brevifolia seeds recruited seedlings in the first year postfire. The advantage of P. fasciculata in greenhouse CFV trials was not replicated in field seedling recruitment. Seedling mortality over the first 3 years postfire was not statistically different among the 3 cache sizes for either species, nor did seedling sizes differ significantly by cache size and survey date. I conclude that, although the overall recruitment from postfire caches was low (12%–17%), prefire caches of short-lived seeds probably recruit postfire seedlings, but only in specific circumstances.