The macroecological abundance-occupancy relationship is well known; not so the potential one of patchiness with either or both. Following earlier work on the intertidal seagrass Zostera capensis in South Africa, interspecific macrofaunal patchiness-occupancy and patchiness-abundance relationships were investigated within each of a number of other seagrass systems (intertidal Cymodocea serratula, Halodule uninervis, Halophila ovalis, Zostera muelleri and Z. noltei, and subtidal Z. capensis) where faunal assemblages were markedly different in their overall abundance and species richness, and where the beds differed in their latitude, longitude, and other variables, including one in an artificial canal in a residential marina. Notwithstanding these differences, in all cases the more abundant and widespread the macrobenthic species, the less was its Lloyd's I patchiness, the more clearly so in respect of occupancy than of abundance. Correlation of I and mean crowding (I ) values was relatively poor, and often not significant. This suggests that patchiness of a species is more influenced by unoccupancy levels than by even marked variation in abundance at occupied sites (e.g. I values of <0.3–>380). Indeed, values of I were closely correlated with the expression [a.unoccupancy + (1−a).I ], where a is >0.80. In all cases, component macrofaunal species displayed a significant or near significant negative patchiness-occupancy relationship in the form of a power-law with a mean scaling coefficient across sites of −0.76, although data points appeared highly scattered. There was little uniformity amongst the component species in the life-style of the most patchy, most widespread or most abundant. p p c c p c