Niche availability, niche overlap and competitive ability are key determinants of the distribution and abundance of species. However, quantifying each of these components is difficult because it is not always possible to identify or measure relevant environmental gradients (niche axes) along which species might partition or compete for niche space. We describe a method that uses seed addition experiments to quantify the number of ‘safe-sites’ (microsites suitable for a species to recruit from seed) at a location and show how this method can be used to quantify niche availability, niche overlap and competitive ability. We illustrate our approach using two seed addition experiments in grassland. In the first experiment, we added seeds of one native and two exotic grass species, alone and in mixture, to plots that were arrayed along a gradient of soil moisture availability. We show that the three species partitioned safe-sites, implying that all three species could locally co-occur through niche partitioning, in part due to different responses to moisture availability. In the second experiment, we added seeds of three commonly co-occurring native grass species, alone and in mixture, to plots with no obvious environmental gradients. One species out-competed two others for site occupancy, allowing us to quantify both the degree of niche overlap and the relative ability of each species to compete for safe-site occupancy without a priori knowledge of the niche axes. Results from both experiments demonstrate the fine-scales at which species can partition niches to facilitate co-occurrence. Synthesis. By conceptualising a plot of ground as containing a limited number of microsites that are safe for the recruitment of a given species, and using seed addition experiments to measure both the number of safe-sites and degree of safe-site overlap among species, we show how niche availability, niche overlap and competitive ability can be quantified at fine-scales without a priori knowledge of niche axes. Our approach allows questions about niche availability and competition for shared niche space to be empirically tested, and to examine how these processes vary along environmental gradients to shape species distributions and patterns of co-occurrence.