Establishment, growth and survival of Eucalyptus regnans F. Muell. seedlings was studied at two sites over three years under a range of small gaps (up to 30mx30m) and large gaps (50mx50m up to clearfells) in a gap cutting experiment in mountain ash forest at Tanjil Bren, Victoria, Australia. In both the group of smaller and larger gaps seedling density at Year 3 was significantly affected by year of treatment, gap size, and/or site. In both groups, densities generally increased with an increase in gap size. Seedling height at Year 3 of the dominant/co-dominant seedlings varied between 2 and 5m, and was positively related to gap size in the group of larger gaps (but not in the group of smaller gaps). In the larger gaps, seedling growth was higher on burnt than on mechanically disturbed seedbeds. Between 20 and 70% of all seedlings survived from Year 1 to Year 3, and survival percentages seemed to increase with increasing gap size. The results of this study indicate that a gap-cutting system is potentially a viable silvicultural treatment in mountain ash forest as most gap sizes would have been considered adequately stocked with E. regnans regeneration three years after logging. However, regeneration success is significantly affected by year-to-year and site variation, and the use of smaller gap sizes may result in regeneration failure in some years and/or at some sites.