Question Is primary dormancy required for seed bank persistence of Leucopogon species? How does the distinction between dormancy and persistence affect our understanding of seed bank dynamics in fire-prone regions?Location: Sclerophyllous plant communities, Sydney region, southeastern Australia.Methods: Seed bank longevity of three morphophysiologically dormant species was ascertained using seed burial trials. Seeds of each species were buried in situ in replicate mesh bags and retrieved annually over three years. Laboratory germination trials and embryo growth measurements over time were used to establish in which order each component of dormancy was overcome, and whether seed banks persisted after the loss of primary dormancy.Results: All species had long-term persistent seed banks, with estimated half-lives between 3.5 and 5.5 years. Physiological dormancy was broken by stratification prior to embryo growth, during the first year of burial. Seeds were able to germinate after retrieval, in light and at warm temperatures. Only low levels of germination occurred in situ over three years of burial, and did not increase over time. During germination trials, embryos remained underdeveloped unless seeds germinated, indicating that seeds were committed to germination once embryo growth was initiated.Conclusions: Long-lived seed banks are an important trait of plants from fire-prone communities, but cues observed to promote germination in situ are not necessarily the ones that break dormancy. A failure to distinguish between these two factors can divert attention away from mechanisms actually controlling dormancy and seed bank dynamics.