Wild vertebrates usually avoid ground disturbed by humans but consequences for their distribution and density are uncertain. The local distribution of capercaillie shifted after an increase in disturbance along woodland tracks adjacent to an expanding Scottish village. We surveyed the birds’ droppings before and after the building of 30 new houses, and model the probability of finding droppings (Pf) in relation to period plus two disturbance gradients – distance to a much disturbed ‘entry zone’ by the village (dE) and ‘distance to nearest track’ (dT). Estimates of Pf are benchmarked to average Pf (Pfav) – a notional scenario in which the birds’ distribution is unaffected by tracks. Change between periods occurred mainly on a strip of ground centred on tracks and averaging 80 m wide, where Pf fell from about 0.5 Pfav before the development to 0.2 Pfav after it. By contrast, Pf on ground 120–260 m from tracks, under a third of the 273 ha main study area, remained at about 3 Pfav throughout the study – indicating a net influx of capercaillie displaced from ground beside tracks in both periods. No capercaillie droppings were found in the entry zone. Beyond this zone, throughout the study, Pf increased as tracks sparsened until dE approached 400 m – whereupon track density and Pf steadied together. Beyond 400 m, Pf remained depressed on ground near tracks (dT ⪅ 100 m). New desire paths after the development caused the proportion of ground where dT < 100 m to increase slightly, from 56% to 60%. Birds on roughly half of a 50 ha refuge should be undisturbed by direct effects of track-based activities – but, if increases in density caused by displaced birds are also deemed disturbance, a refuge would need to be over 3 km2 to keep half of it undisturbed.