How individuals respond to habitat heterogeneity is usually measured as variation in range size and by ranking the relative importance of habitat types (habitat selection). The combined effect of how individuals incorporate different habitat types in their home ranges and allocate their time budget between them is rarely derived. Additionally, when home range size varies between individuals, habitat selection analyses might be flawed if foraging decisions are based on variation in absolute rather than proportional availability. We investigated the suitability of standard analytical approaches by measuring the spatial responses of tawny owls to habitat fragmentation. These owls inhabited woodland of various sizes, representing a fragmentation gradient from open farmland with small, isolated woodland patches, to continuous woodland within their home ranges. In 17 territories within open farmland, the available area covered by woodland increased with the square root of the area of open land embraced in the home range. The owls did not display functional response in habitat selection, but females selected woodland more strongly than males. Females utilised woodland 10 times more intensively in farmland than in continuous woods, whereas males utilised farmland woods 3.2 times more intensively. Moreover, females in farmland exploited woodland 3.2 times as intensively as males, apparently because of higher travel costs in open areas. Since the extensive variation in intensity of use as a function of total availability was not indicated from the analysis of habitat selection, we suggest that information about intensity of use be more widely used as a supplementary measure of habitat use patterns than appears to be the practice at present.