Pollinators have long been implicated in plant speciation. Peter Raven's earlier work was instrumental in integrating foraging energetics of animals into our understanding of how shifts in major pollinators influence the evolutionary diversification of floral traits. More recently, efforts by Raven and others in the area of conservation have inspired pollination biologists to consider the implications of pollinator shifts and losses due to human activities. This paper uses the shift between hummingbird and hawkmoth pollination as a model for exploring impacts of pollinator shifts on plant populations. Recent studies have quantified the degree of reproductive isolation due to such pollinators in several genera. Data from Ipomopsis Michx. further allow us to consider whether recent changes in pollinator regimes have demographic consequences for plant populations. A majority of plant populations may currently suffer from pollen limitations on seed production, but few data exist on the demographic consequences of poor reproduction. In Ipomopsis, reduced seed production due to pollen limitation can impact the number of individuals surviving to reproduce in the next generation. Some populations of I. tenuituba (Rydb.) V. E. Grant are estimated to have finite rates of increase less than unity, which can be explained in part by current low levels of hawkmoth pollination. In the absence of an increase in hawkmoths, selection for wider corolla tubes and other floral traits could, in principle, attract enough hummingbird pollination to result in a growing population, but models show that such evolution by natural selection may leave the population vulnerable to local extinction. We need more studies of the quantitative demographic consequences of changes in pollinator regimes. Such studies should consider how evolutionary changes influence the risk of extinction.