In urban areas the density of breeding tits (Paridae) is generally higher and reproductive performance lower compared to rural areas. To explain these landscape differences several hypotheses have been proposed, e.g. differences in habitat quality, inter- and intra-specific competition, predation and food abundance. How breeding performance of birds within remnants of natural vegetation in urban areas is affected by adjacent matrix has been less studied. We performed an experimental study in four urban woodlands surrounded by three types of habitat matrix: residential, high-rise building and grassland/golf course. We placed 300 nest boxes for tits (great tit Parus major and blue tit Cyanistes caeruleus) in 15 transects that were 200 m long, extending from the adjacent matrix 150 m into urban woodland. Density of breeding great tit pairs was highest in residential areas and hatch date was earlier in the residential habitat compared to the other two habitats, however, nestling condition was lower in residential areas. Hatching date was earlier but hatching spread (heaviest nestling/lightest nestling) higher in the three types of urban matrix than inside the urban woodlands. In contrast to previous large-scale comparisons of urban and rural matrix, we almost exclusively found differences in qualitative measures (nestling condition and hatchling spread) at the small scale in which this study was conducted. Adjacent matrix affect great tits breeding in remnant urban woodlands, thus we suggest that management of the surrounding matrix should be included in conservation plans for urban woodlands.