Invasive species distributions tend to be biased towards some habitats compared to others due to the combined effects of habitat-specific resistance to invasion and non-uniform propagule pressure. These two factors may also interact, with habitat resistance varying as a function of propagule supply rate. Recruitment experiments, in which the number of individuals recruiting into a population is measured under different propagule supply rates, can help us understand these interactions and quantify habitat resistance to invasion while controlling for variation in propagule supply rate. Here, we constructed recruitment functions for the invasive herb Hieracium lepidulum by sowing seeds at five different densities into six different habitat types in New Zealand's Southern Alps repeated over two successive years, and monitored seedling recruitment and survival over a four year period. We fitted recruitment functions that allowed us to estimate the total number of safe sites available for plants to occupy, which we used as a measure of invasion resistance, and tested several hypotheses concerning how invasion resistance differed among habitats and over time. We found significant differences in levels of H. lepidulum recruitment among habitats, which did not match the species' current distribution in the landscape. Local biotic and abiotic characteristics helped explain some of the between-habitat variation, with vascular plant species richness, vascular plant cover, and light availability, all positively correlated with the number of safe sites for recruitment. Resistance also varied over time however, with cohorts sown in successive years showing different levels of recruitment in some habitats but not others. These results show that recruitment functions can be used to quantify habitat resistance to invasion and to identify potential mechanisms of invasion resistance.