Abstract Three treatments designed to initiate the process of restoring the surface fire regime and open forest structure of a southwestern ponderosa pine forest were compared on the Kaibab National Forest along the Grand Canyon’s South Rim. The treatments were: (1) full restoration (FULL)—thinning trees to emulate stand structure prior to fire regime disruption ca. 1887, forest floor fuel treatment, and prescribed burning, (2) minimal thinning (MIN)—removing young trees only around living old-growth (pre-1887) trees, fuel treatment, and prescribed burning, (3) burn-only (BURN)—representing the current management policy in Grand Canyon National Park (GCNP), and (4) CONTROL. Each treatment was applied to a 12 ha unit. Compared to reconstructed 1887 conditions, all study sites were much more dense prior to treatment (94–176 trees/ha in 1887, compared to 783–3693 trees/ha in 1997). However, basal area increases were less striking (12.6–20.3 in 1887, 17.5–27.0 m 2/ha in 1997), reflecting past harvest and dwarf mistletoe reduction treatments that removed many large pines. In 2000, 1 year after treatment, tree densities were reduced to 11, 23, and 37 of pre-treatment levels in the FULL, MIN, and BURN treatments, respectively. Understory plant communities showed significant declines in richness and plant frequency across years, probably due to a severe drought in 2000 (60% of average precipitation). No differences in plant communities were observed across treatments, despite the mechanized disturbance associated with tree removal in the FULL treatment. Prescribed fire behavior (flame length, flaming zone depth) and effects (bole char, crown scorch) were similar across all three burned treatments. Simulated fire behavior under dry, windy conditions was reduced in all three treatments compared to the control. The FULL treatment was much less susceptible to crownfire due to reduced crown bulk density and crown fuel load and increased crown base height. Crownfire susceptibility of the BURN treatment was only slightly reduced, while the MIN treatment was intermediate. Compared to the reference conditions of forest structure, the FULL treatment represented the most rapid and comprehensive restoration treatment, although the residual stand was at the low end of historical density. The BURN treatment thinned many small trees but had minor effects on crownfire susceptibility. Effects of the MIN treatment fell between FULL and BURN. The experimental treatments may be useful for the creation of defensible firebreaks near developments, roads, and boundaries with the FULL treatment, supplemented by MIN and BURN treatments over larger areas.