Browsing of mountain beech seedlings by introduced deer in the central North Island of New Zealand appears to have inhibited canopy regeneration over large areas. In 1998, a trial of high-, medium- and low-intensity deer-culling treatments was initiated in Kaimanawa and Kaweka Forest Parks to test whether mountain beech (Nothofagus solandri var. cliffortioides) forest regeneration could be restored by removing deer. Paired exclosure plots (one plot fenced to exclude deer and the other unfenced) were established within a high-intensity culling area, to monitor the benefits of recreational, commercial and aerial deer culling. Paired plots were also established within low- and medium-intensity culling areas. Medium-intensity culling was allowed through recreational and commercial deer culling. In two low-intensity culling treatment areas, deer management remained substantially unaltered. Annual relative growth rates of tagged seedlings from spring 1998 to spring 2001 from low-, medium- and high-intensity culling areas provide strong evidence that mountain beech seedling growth increases once browsing by deer is removed through fencing. Faecal pellet data indicated that high-intensity deer culling reduced deer abundance by 67% in comparison to medium- and low-intensity culling areas. This apparent reduction in deer abundance appears to have led to a doubling in mountain beech seedling growth in the high-intensity culling area outside fences, in comparison to low- and medium-intensity deer culling areas where there was little or no evidence of benefits for seedling growth.