We used individual colour-marking and territory mapping to quantify the spatial extent of 32 Marsh Tit Poecile palustris territories in Monks Wood, eastern England, during spring of the years 2002–04. A mean territory size of 4.1 ha was identified. All 2004 spring territories were located, allowing a breeding density of 14 pairs/km2 to be calculated. Availability of airborne digital remote sensing (LiDAR) data for Monks Wood allowed the characterization of the canopy structure in territories and non-breeding areas using a three-dimensional canopy-height model. The difference between the mean canopy height of the 2004 territories and that of the unoccupied area of the study site in the same year was 1.8 m, or 14%. Sampling the unoccupied area, with hypothetical ‘pseudo-territories’, showed a statistically significant difference of 1.6 m (13%) between the mean canopy heights of the ‘taller’ 2004 territories and the unoccupied pseudo-territories. A comparison by field survey of tree and shrub species composition between the 2004 territories and pseudo-territories found no difference in species richness or the mean density of shrubs or mature trees (> 30 cm diameter at breast height, dbh). The mean density of medium-sized (5–30 cm dbh) and small (< 5 cm dbh) trees was, respectively, 1.9 and 3.9 times greater in the pseudo-territories, values that were statistically significant. Overall, Marsh Tits in Monks Wood appeared to require mature trees with a shrub layer beneath the top canopy, but avoided areas with large numbers of young and immature trees.