© The Author(s), 2016. This is the author's version of the work and is distributed under the terms of the Creative Commons Attribution License. The definitive version was published in Current Biology 27 (2017): 149-154, doi:10.1016/j.cub.2016.10.053. ; The degree to which offspring remain near their parents or disperse widely is critical for understanding population dynamics, evolution, and biogeography, and for designing conservation actions. In the ocean, most estimates suggesting short-distance dispersal are based on direct ecological observations of dispersing individuals, while indirect evolutionary estimates often suggest substantially greater homogeneity among populations. Reconciling these two approaches and their seemingly competing perspectives on dispersal has been a major challenge. However, here we show for the first time that evolutionary and ecological measures of larval dispersal can closely agree by using both to estimate the distribution of dispersal distances. In orange clownfish (Amphiprion percula) populations in Kimbe Bay, Papua New Guinea, we found that evolutionary dispersal kernels were 17 [95% CI: 12–24] km wide, while an exhaustive set of direct larval dispersal observations suggested kernel widths of 27 [19–36] km or 19 [15–27] km across two years. The similarity between these two approaches suggests that ecological and evolutionary dispersal kernels can be equivalent, and that the apparent disagreement between direct and indirect measurements can be overcome. Our results suggest that carefully applied evolutionary methods, which are often less expensive, can be broadly relevant for understanding ecological dispersal across the tree of life. ; Funding was provided by an NSF graduate fellowship, an NDSEG graduate fellowship, an International Society for Reef Studies fellowship, an Alfred P. Sloan Research Fellowship, NSF OCE-1430218, and the King Abdullah University of Science and Technology. ; 2017-12-15