Abstract Rapid infiltration influences rainfall partitioning, subsurface water movement, and freshwater availability, yet its large‐scale spatial organization remains poorly constrained by limited observations and the lack of scalable approaches for synthesizing infiltration measurements. Here, we identify rapid infiltration signatures from a global database using a dual‐mechanism modeling framework. Rapid infiltration signatures occurred in 26.6% of 4,992 valid curves. Occurrence varied among measurement methods, while rapid and non‐rapid signatures frequently coexisted within closely matched local contexts. At identical recorded locations, 74.7% of qualifying groups contained both rapid and non‐rapid signatures; for rapid‐classified curves, the fitted rapid component accounted for a median 41.2% of cumulative infiltration. Across 83 sites, rapid‐signature odds decreased by approximately 26% per 10° increase in absolute latitude. These findings provide a large‐scale observational basis for investigating rapid infiltration and highlight the importance of measurement context and independent physical evidence for improving its representation in hydrological and Earth system models.