We used five-minute bird counts to investigate whether introduced Australian magpies (Gymnorhina tibicen) influence the abundance of other birds in rural New Zealand. Over 3 years, magpies were removed from five c. 900-ha study blocks, one in each of Northland, Waikato, Bay of Plenty, Wellington and Southland. Birds were counted in both the treatment blocks and paired non-treatment blocks for the 3 years of removal and also 1 year before. To minimise problems raised elsewhere with index counts we (1) selected treatment blocks and count stations using randomisation procedures, (2) used trained observers who spent equal time in paired treatment and non-treatment blocks, and (3) counted all blocks at the same time of year and only in good weather. On average, 548 magpies were removed from each treatment block each year, with magpie counts reduced by 76% relative to non-treatment blocks. Our results suggest magpies may restrict the movements of some birds (including kererū and tūī) in rural areas, but are less important than pest mammals at limiting population abundance at a landscape scale. We submit that five-minute bird counts were appropriate for our objectives, but that more research to examine their relationship to absolute densities is needed.