As important marine mesopredators and sensitive indicators of Antarctic ecosystem change, penguins have been a major focus of long-term biological research in the Antarctic. However, the vast majority of such studies have been constrained by logistics and relate mostly to the temporal dynamics of individual breeding populations from which regional trends have been inferred, often without regard for the complex spatial heterogeneity of population processes and the underlying environmental conditions. Integrating diverse census data from 70 breeding sites across 31 years in a robust, hierarchical analysis, we find that trends from intensely studied populations may poorly reflect regional dynamics and confuse interpretation of environmental drivers. Results from integrated analyses confirm that Pygoscelis adeliae (Adélie Penguins) are decreasing at almost all locations on the Antarctic Peninsula. Results also resolve previously contradictory studies and unambiguously establish that P. antarctica (Chinstrap Penguins), thought to benefit from decreasing sea ice, are instead declining regionally. In contrast, another open-water species, P. papua (Gentoo Penguin), is increasing in abundance and expanding southward. These disparate population trends accord with recent mechanistic hypotheses of biological change in the Southern Ocean and highlight limitations of the influential but oversimplified “sea ice” hypothesis. Aggregating population data at the regional scale also allows us to quantify rates of regional population change in a way not previously possible.