Annual estimates of population size are important for tracking long-term population trajectories. There is concern that monarch butterflies (Danaus plexippus L.) in eastern North America are declining because the Mexican overwintering colonies have been shrinking over the past 20 yr. We examined a 19-yr data set of fall monarch censuses at a site in northern Michigan (Peninsula Point) to determine if the number of monarchs counted here has changed over time. At this site, volunteers walk a standardized transect daily during the fall to record migratory monarchs. Our secondary goal was to evaluate the relationship between the annual numbers at Peninsula Point and the size of the subsequent overwintering colonies. A statistical model that included multiple predictors of 1,793 census counts revealed a small but significant positive effect of year on counts, but the mean annual count number was not significantly correlated with year. In either case, we found no evidence for declines in monarch abundance. We also found no relationship between annual estimates of abundance at Peninsula Point and annual estimates of overwintering colony size. We interpret the disparity between long-term patterns here versus those in Mexico to be the result of mortality during migration or other factors that reduce migration success. Given this conclusion, we argue that preserving migratory habitats and resources, as well as reducing anthropogenic activities that hinder migration, should be the highest priorities for conserving monarchs in eastern North America.