Most habitat complexity–species diversity relationships are specific to the habitat type, taxonomic group and spatial scale investigated. There is a need for a biologically relevant metric of habitat complexity that is transferable across systems and scales. Here, we introduce a novel metric of landscape microhabitat complexity that may overcome such constraints. It is scaled to the size of individual organisms and it numerically captures structural elements of the habitat that are biologically relevant to individuals. Our model ecosystem to develop this metric was a rocky shore in Eastern Australia. Here species richness of invertebrates was linked to habitat complexity quantified by the new metric. Importantly, measurements made at the scale of the organism predicted species richness at the landscape scale. Testing the generality of this approach across different systems may prove useful in a broader conservation biology context.