Related species of similar morphology can differ greatly in distribution and abundance. Elucidating reasons for such differences can contribute to an understanding of intrinsic limiting factors and the causes of rarity. We studied sympatric populations of two terrestrial lilies with contrasting distributions: Calochortus lyallii, which is geographically restricted but locally abundant, and C. macrocarpus, which is widespread but locally sparse. Marked plants of each species were monitored for 5 years in British Columbia, Canada. Matrix projection models were used to estimate annual and stochastic population growth rates (λ and λs) and to compare demographic traits. Annual λ-values ranged from 0.89 to 1.04 in C. lyallii and from 0.89 to 1.01 in C. macrocarpus. Stochastic projections yielded a long-term growth rate near 1 for C. lyallii, but indicated a decline for C. macrocarpus. Elasticity analysis indicated that over the 5-year period of the study, survival of flowering plants made a larger proportional contribution to λ in C. lyallii than in C. macrocarpus. LTRE analysis showed that temporal variation in λ was driven primarily by the dynamics of flowering individuals in C. lyallii, and by the dynamics of vegetative individuals in C. macrocarpus. Similarly, higher flowering rates in C. lyallii and greater vegetative stasis in C. macrocarpus made the largest contribution to the difference in λ between species. Thus, local persistence in these two morphologically similar species appears to be achieved via different demographic pathways. Our analyses show that extrapolations about demographic processes and population dynamics based on taxonomic relatedness, morphological similarity or habitat overlap may often not be justified.