During the springs of 1995-1997, we studied birds and landscapes at 70 sites in the Chihuahuan Desert to assess relations between bird community structure and landscape patchiness. Within each of two spatial extents (1-kin and 2-km-radius areas centered on each site), we measured the number of patches of individual land-cover types and the total number of patches of all land-cover types. Mean bird richness, and the mean abundance and probability of occurrence of most bird species were significantly correlated with one or more of these variables. Contrary to evidence from other systems, positive association with landscape patchiness did not increase with the degree to which species were habitat generalists, was not negatively related to body size, and did not differ between neotropical migrants and nonmigrants. For the communities' primary constituent species as a group, the strength of positive and negative associations with patchiness did not differ between landscape extents. Within the 1-km but not the 2-km extent, habitat specialists were more positively and negatively associated with patchiness than were habitat generalists. In general, however, neither habitat breadth, body size, nor migratory status seemed to be responsible for associations with landscape patchiness. Mean richness, and the mean abundance and probability of occurrence of most species were significantly correlated with patchiness within one or both extents, and patchiness of all of the most extensive land-cover types was influential. The simplest explanation for most of the bird-patchiness relations we found is that the associations reflected species-specific habitat needs. Through effects on avian richness, abundance, and occurrence, landscape patchiness affected bird community structure. A more complete understanding of the effects of landscape patchiness on bird community structure is likely to emerge when ecologists study the patchiness of major land-cover types at various spatial extents.