Fighting and accidental injury commonly cause black rhinoceros (rhino; Diceros bicornis) death after release. Smaller reserves and higher conspecific density after release (release density) might increase a rhino's encounter rate with hazards like fenced boundaries and conspecifics. We conducted a science-by-management experiment on the influence of reserve size and release density on rates of movement, association, and injury and death amongst 39 black rhinos during the first 100 days after their release into 4 Namibian and 8 South African reserves ranging in size from 670 ha to 45,000 ha. Association rates were negatively related to reserve size and positively correlated with release density. There was also a negative relationship between the proportion of the reserve traversed by individual rhinos and reserve size. In reserves ≥18,000 ha association rates were consistently zero but became elevated in reserves ≤11,500 ha and at release densities ≤9 km²/rhino. Daily displacement did not increase with increasing reserve size >8,500 ha but in smaller reserves daily displacements indicated higher encounter rates by released rhinos with fenced boundaries. Three rhinos received fight-related injuries requiring intervention and 2 of 4 deaths were fight-related. All injuries and 3 deaths occurred in reserves ≤11,500 ha. Model selection based on Akaike's second-order Information Criterion indicated that the parameter release density alone best explained mortality risk. Traditionally considered risk factors, rhino sex, age, and presence of resident conspecifics, were superseded by the risk posed by releases into smaller reserves. Reserves ≤11,500 ha and release densities ≤9 km²/rhino pose an increasing risk to rhino survivorship and so larger reserves and lower densities than these should be favored as release sites.