Aim: Long-distance dispersal (LDD) events occur rarely but play a fundamental role in shaping species biogeography. Lying at the heart of island biogeography theory, LDD relies on unusual events to facilitate colonization of new habitats and range expansion. Despite the importance of LDD, it is inherently difficult to quantify due to the rarity of such events. We estimate the probability of LDD of the seagrass Heterozostera nigricaulis, a common Australian species, across the Pacific Ocean to colonize South America. Location: Coastal Chile, Australia and the Pacific Ocean. Methods: Genetic analyses of H. nigricaulis collected from Chile and Australia were used to assess the relationship between the populations and levels of clonality. Ocean surface current models were used to predict the probability of propagules dispersing from south-east Australia to central Chile and shipping data used to determine the likelihood of anthropogenic dispersal. Results: Our study infers that the seagrass H. nigricaulis dispersed from Australia across the entire width of the Pacific (c. 14,000 km) to colonize South America on two occasions. Genetic analyses reveal that these events led to two large isolated clones, one of which covers a combined area of 3.47 km2. Oceanographic models estimate the arrival probability of a dispersal propagule within 3 years to be at most 0.00264%. Early shipping provides a potential alternative dispersal vector, yet few ships sailed from SE Australia to Chile prior to the first recording of H. nigricaulis and the lack of more recent and ongoing introductions demonstrate the rarity of such dispersal. Main conclusions: These findings demonstrate LDD does occur over extreme distances despite very low probabilities. The large number of propagules (100s of millions) produced over 100s of years suggests that the arrival of propagules in Chile was inevitable and confirms the importance of LDD for species distributions and community ecology.