Many bee populations are declining globally, in large part due to habitat loss and the resulting nutritional shortages. Tallgrass prairie restoration efforts are often aimed at establishing vegetation that resembles undegraded historic conditions and to promote pollinator habitat. Restored prairies in southern Wisconsin have been shown to have increased floral resource abundance and to support a greater abundance and diversity of bumble bees than non-restored old agricultural fields. However, floral resource abundance does not appear to be the only driver of bumble bee abundance. Many factors have been shown to affect bumble bee foraging patterns: flower color, floral resource abundance, nectar availability, and the quality of pollen. Recent research has shown that, in experimental settings, bumble bees prefer pollen with a macronutrient composition that has a high protein:lipid ratio (P:L), suggesting that bumble bee larvae development and survival depends on specific pollen requirements. For this reason, we expect nutritional quality may be a potential mechanism for variation in local bumble bee abundance and diversity in restored prairies. However, the nutritional quality of pollen is typically unknown for most flowering species found in tallgrass prairies. Acquiring the pollen macronutrient composition of flowering species found in prairies would allow us to estimate the community-level nutritional quality of restored prairies. Here we ask, do restored prairies with high community-level P:L pollen have more abundant and diverse bumble bee populations than sites with a low community P:L pollen? To answer this question, we collected floral resource and bumble bee data during the summers of 2019-2022. Pollen from common herbaceous prairie species was collected and macronutrient assays were completed in 2021 and 2022. We conducted protein and lipid assays of pollen from 80 common flowering species in tallgrass prairies. We then calculated the community-level pollen P:L by multiplying the P:L of pollen for each flowering species found at a site by the flowering species abundance at that site and totaling all species at that site. Using preliminary data, we compared sites where we had over 80% of P:L ratios for the species observed (n=20; p-value: 0.056; r-squared: 0.25). A linear model with pollen nutritional quality and floral resource abundance explained 53% of variation in bumble bee abundance (p-value: 0.01). Our data support our hypothesis that pollen macronutrient composition drives bumble bee abundance, however it does not appear to affect bumble bee diversity.