We investigated the effects of food supply on decisions made by dispersing juvenile Tawny Owls Strix aluco in Kielder Forest, Northumberland, in 1996 and 1997. Field Voles Microtus agrestis were the main food of the owls and clear-cuts the main habitat for voles. A vole sign index was used to estimate vole abundance. In areas near to roosting owls, mean vole densities were 83 and 115 ha−1 in 1996 and 1997, respectively. The prediction that birds would perform area-restricted searches when prey was more abundant was not confirmed. Moreover, we found no evidence that juveniles avoided conspecifics. Owls appeared to have an imperfect knowledge of the environment as they responded to variability in Field Vole densities by altering the time spent in different areas rather than by moving to areas with successively greater vole densities. Vole abundance explained 25.7% of the variation in the time spent in different areas. Movements did not decrease with time after dispersal, although the detection of such movements was prone to error. This study supports recent work suggesting that although dispersal may be initiated by a variety of proximate and ultimate factors, individual decisions made during dispersal may depend partly on environmental conditions encountered during the process itself.