Species range shifts often speed ahead of their modeled climatic niches.

Anticipating how species distributions will shift with climate change is key for biodiversity conservation and management. Commonly, species' range shifts are observed by analyzing changes in occurrence or abundance data through time, or predicted across different climate change scenarios by modeling species' climatic niches. However, it remains unclear how well these climate-based forecasts align with empirically documented range shifts from monitoring efforts. Here, we tested the congruence between modeled range shifts, predicted using climatic niche models, and documented range shifts, derived from empirical observations collected over recent decades, for more than 9,500 range shifts across over 3,500 marine and terrestrial species. We found that documented and modeled range shifts tend to align in latitudinal direction, with greater alignment for marine (76%) than terrestrial (56%) cases. However, even when the directions aligned, documented shifts exceeded modeled shifts in 62% of cases, nearly twice as often as they lagged behind (38%), and their median rates were four times faster than those of the modeled shifts. Our findings suggest that climate-based models can approximate observed range dynamics under specific conditions, particularly over long time periods and restricted spatial areas, when habitats remain well connected and under low climate fluctuations over time. These insights provide valuable guidance for both improving predictions and informing responses to climate-driven biodiversity redistribution.