Most studies report the abundance of plastic items in the environment, but mass is an equally important currency for monitoring plastic pollution, particularly given attempts to balance the global plastic budget. We determined the size/mass composition of litter stranded on a remote, infrequently-cleaned sandy beach on the west coast of South Africa. Traditional surveys of superficial macro-litter were augmented by sieved transects for buried macro-litter (8-mm mesh), meso-litter (2-mm mesh) and sediment cores for micro-litter. Aggregating the data across all sampling scales, the total density was ∼1.9 × 10 5 anthropogenic particulate pollutants per linear meter of beach, 99.7% of which were microfibers (most of which are likely not 'plastic'). Plastics comprised 99.6% of beach macro-and meso-litter by number and 89% by mass. Small items dominated samples numerically, but were trivial relative to larger items in terms of their mass. Buried litter accounted for 86% of macro-plastic items, but only 5% of the mass of macro-plastics, because smaller items are buried more easily than large items. The total density of plastic (∼1.2 kg•m −1), at least half of which was from fisheries and shipping, is much lower than predicted by global models of plastic leakage from land-based sources. Ongoing degradation of plastic items already in the environment, particularly on beaches, is likely to result in a marked increase in plastic fragments, even if we stop leaking additional plastic. The collection of large items from beaches is a useful stopgap measure to limit the formation of micro-plastics while we formulate effective steps to prevent plastic leakage into the environment.