Mutualisms may facilitate coupled population expansion or decline if changes in either partner species lead to similar changes in the other species. However, few studies have measured how both species in a mutualism respond to changes in each other's abundances. In this study, I tested how reproduction of two mutualists—the bumble bee pollinator, Bombus appositus , and the perennial wildflower, Delphinium barbeyi—responded to natural and experimental changes in partner density and the resources they derive from each other (food for bees and pollen transfer for plants). From the bumble bee's perspective, bumble bee foragers flew short distances (< 100 m), suggesting that forager density was not strongly limited by floral resource availability or that long-distance flights were costly. Only early forager densities (founding queens) were higher in areas with naturally more flowers, but mid- and late-season forager densities (workers and males) did not increase in areas with more flowers. Bee reproduction was constant across meadows that varied naturally in flower availability, and nectar in D. barbeyi flowers was only depleted by 28% throughout the day during the period of colony growth. However, experimental food addition generally increased gyne production. From the plant's perspective, in natural D. barbeyi patches, pollen receipt increased linearly with density, but this effect was only significant in meadows and not in forests. Seed production increased across lower flower densities and decreased across higher flower densities, but this effect was only statistically significant in meadows. Experimental flower density did not affect pollinator visitation rates, pollen receipt, or seed production. Neither natural increases in pollinator visitation rate nor supplemental hand-pollinations affected pollen receipt or seed production. Bumble bee pollinator exclusion generally reduced seed set, but sometimes fly pollinators (which have seed predator offspring) compensated for lost bumble bee pollination services. Together, these results suggest that in this mutualism, neither species should respond rapidly to small changes in partner abundance.